Post on 23-Sep-2020
TOWARDS DIUX 2.1 OR 3.0?
EXAMINING DEFENSE INNOVATION UNIT EXPERIMENTAL’S PROGRESS TOWARDS
PROCUREMENT INNOVATION
A THESIS
SUBMITTED TO THE
INTERSCHOOL HONORS PROGRAM IN INTERNATIONAL SECURITY STUDIES
Center for International Security and Cooperation
Freeman Spogli Institute for International Studies
STANFORD UNIVERSITY
By
Gabriele Fisher
June 2017
Adviser:
Dr. Herb Lin
Fisher
!ii
Fisher
Abstract
Defense contracting is responsible for procuring the tools used by the defense community. Defense procurement, however, has become emblematic of scheduling delays and cost overruns, fueled by the outsized influence of traditional Beltway military contractors in defense acquisition. As American adversaries utilize novel technologies across domains physical and cyber, the DoD faces pressure to match this pace of innovation in their defense strategies. In 2015, then-Secretary of Defense Ashton Carter announced the creation of Defense Innovation Unit Experimental (DIUx), an outfit within the DoD to expedite contracting with small technology outfits pioneering defensive technologies. DIUx has faced criticism for not addressing the federal acquisition regulations that render the DoD a difficult customer for small technology outfits, in addition to challenges in continued engagement with defense startups. This thesis examines the realities underlying these criticisms. Specifically, it argues that DIUx occupies a previously missing junction in the defense innovation ecosystem: namely, the process of connecting DoD-supported R&D to defense community customers and facilitating contracts between these two entities. It also finds that DIUx has made marked improvements in facilitating relationships with small technology firms by utilizing Other Transaction Authorities (OTAs) to expedite contracting. OTA contracts, however, may not be conducive to longer-term business relationships between the DoD and small technology vendors. The adoption of OTAs raises further questions about DIUx’s balance between serving as a federal acquisition regulation reformer versus serving as a relationship manager between the DoD and high-technology firms.
!iii
Fisher
!iv
Fisher
Acknowledgements
Without the support of my mentors and friends, this thesis would not exist. Thank you for supporting me through the lows of late-night chapter drafting and the highs of presenting my research. You know your relationships with people are meaningful if they are enthusiastic to hear you talk about defense contracting, whose technicality and bureaucratic nature strike fatigue and usually not excitement!
I’m especially indebted to my advisor Dr. Herb Lin for supporting my research, providing sharp constructive criticism with kindness, and serving as one of my mentors as I scope out my path in the defense technology world. Without an inspiration like Herb, I’m less sure that this engineer would have branched out beyond computer science lecture halls to consider the ramifications and utility of her work. Thank you for your enthusiasm, endless reading recommendations, and the time you invested in my research. I’ll never forget your offer to go over my thesis presentation slides over Skype at nine PM the eve of my presentation, and your patience in going over every single slide. I’m forever grateful for your mentorship and friendship, and our Yank Sing dim sum celebration will have to come soon. I also thank Professors Martha Crenshaw and Chip Blacker for creating an interdisciplinary space for students, hailing from all walks of Stanford life, to grapple with national security issues. Additional thanks to Shiri Krebs and Jia-Qing Zhao for supporting CISAC and the success of its students. Special thanks to Jafer Ahmad at DIUx for helping me contact interviewees, and the DIUx stakeholders willing to talk to me for my research. Finally, I owe infinite thanks to friends and loved ones who supported me through this stressful and hectic senior year. Thanks to my fellow CISAC honors students for empathizing with these challenges and keeping my chin up. I’m also so thankful for the close friends who bolstered me throughout the research and writing of this thesis. Of course, I would be amiss if I did not thank my family for expressing enthusiasm over and interest in my research. I hope these following pages are worth your while.
!v
Fisher
Table of Contents
Abstract iii
Acknowledgements v
List of Acronyms viii
Chapter I | Introduction 1
The Kunduz Strike and its Unexpected Connection to Defense Procurement 3
Why the DoD’s Inclusion of Technology Startups Matters 7
Methodology and Overview 10
Chapter II | Surveying the Defense Innovation Landscape Before DIUx 13
A Brief Overview of In-Q-Tel and DARPA 14
Defense GVC Programs 15
I. Improving Returns on Investment and Returns on Technology 16
II. Stimulating Competition and Innovation in Defense Products 19
Defense Work Programs 22
I. Cultivating R&D to Advance Defense Technologies 23
II. Procuring Advanced Technologies from R&D Supported by Work Programs 25
Federal Acquisition Regulations 27
I. Discouraging Contract Breaches 28
II. Stimulating Competition Among Contractors 30
III. Supporting Small Businesses to Capture New Technologies 32
Rapid Acquisition Task Forces Within the DoD 36
I. Rapid Acquisition Favors the Establishment 36
II. Siloing Expedited Procurement Hampers New Technology Adoption 37
A Summary of Continuing Challenges in the Defense Innovation Ecosystem 38
Chapter III | DIUx’s First Year, its Early Challenges, and its Reinvention 41
DIUx 1.0: Fumbling Towards a Clear Mission 42
I. DIUx 1.0 Management Lacking “Silicon Valley” Traits 42
II. DIUx 1.0 Still Required Expansion of Defense Agency Partners 44
III. DIUx 1.0 Target Companies Too Constrained 44
IV. DIUx 1.0 Strategy Restrained to Performing Assists 45
!vi
Fisher
Introducing DIUx 2.0: A Movement Towards Mission Success 48
I. DIUx 2.0 Management Folds in Veteran Entrepreneurs 49
II. DIUx 2.0 Still Required Expansion of Defense Agency Partners 50
III. DIUx 2.0 Target Companies More Diverse 51
IV. DIUx 2.0 Strategy Utilizes Nontradtional Contracting Vehicles 52
A Summary of DIUx’s Evolution and a Question: What Lies Beyond the OTA? 55
Chapter IV | Case Studies on DIUx 2.0 57
Perspectives from Two DIUx-Contracted Firms 58
I. Shield AI 58
II. Software Vendor 60
Perspectives from DIUx Portfolio Managers and Contracting Facilitators 65
I. Building DIUx Relationships through OTAs 65
II. Skepticism Towards OTA Reliance 68
Observations from the Drafting of a DIUx Problem Statement 70
Conclusions 72
Chapter V | Conclusion: Towards DIUx 2.1, Built on the Success of OTAs 75
References 81
Written Resources 81
Interviews 87
Appendices 88
Appendix A: Federal Acquisition Regulations, 1971-2013 88
Appendix B: DoD Entities Involved in the Fulfillment of Urgent Needs 89
Appendix C: Q4 2016 DIUx Contracts 90
!vii
Fisher
List of Acronyms
!viii
CSO Commercial Solutions Opening
DARPA Defense Advanced Projects Research Agency
DCGS Distributed Common Ground System
DISA Defense Information Systems Agency
DIUx Defense Innovation Unit Experimental
DoD Department of Defense
FAR Federal Acquisition Regulation
GVC Government Venture Capital
IDC Indefinite Delivery Contract
IED Improvised Explosive Device
IP Intellectual Property
JIDO Joint Improvised-Threat Defeat Organization
MRAP Mine-Resistant Ambush Protected [Vehicle]
NDAA National Defense Authorization Act
OSD Office of the Secretary of Defense
OTA Other Transaction Authority
QIC In-Q-Tel Interface Center
SBA Small Business Administration
UAS Unmanned Aerial System
Fisher
Chapter I | Introduction
The day of October 3rd, 2015 was supposed to mark a decisive blow against the Taliban
in Kunduz, the largest city in northwestern Afghanistan. Instead, the day’s conclusion brought
the killing of civilians and medical personnel in a Doctors Without Borders hospital.
A few days prior, the Taliban had staged a coordinated attack on Kunduz, prompting local
Afghan forces to withdraw and leave the city in control of the insurgency. The need to retake
Kunduz from the Taliban was imperative; the fifth most populous city in Afghanistan, it serves as
an agricultural breadbasket for the country. American Special Operations Forces planned an 1
October 3 aerial attack on a newly established Taliban base in Kunduz to scuttle the insurgents’
presence. That night, in black skies, an AC-130 gunship patrolled the airspace above the city,
evading a surface-to-air missile while attempting to locate the Taliban-occupied building. 2
Forced off course by Taliban ground defenses, the gunship and its multiple readied cannons
veered eight miles from their flight trajectory. The Forward Operating Base dispatched live 3
intelligence to the plane to help it return to course. After the pilots fed updated target coordinates
into their fire control systems, these new coordinates brought the plane to an empty field as
opposed to an insurgent base. The information furnished by the base was inaccurate. 4
BBC Newswire. 2016. Kunduz hospital bombing: 16 US forces 'disciplined'. BBC. April 28, 2016.1
Tucker, Patrick. 2016. The war over soon-to-be-outdated army Intelligence Systems. Defense One. July 2
5, 2016. Campbell, Gen John. 2015. Department of Defense Press Briefing by General Campbell via 3
Teleconference from Afghanistan. Department of Defense Press Operations. November 25, 2015. Campbell, 2015.4
!1
Fisher
1000 feet away from the plane, however, a cluster of buildings showed furtive signs of
activity. The aircrew believed the building’s appearance fit the physical description of the 5
targeted Taliban headquarters, and at night, the aircrew were unable to identify any signs
indicating the building’s protected status. The mission commander, seeking to complete mission
objectives while accelerating the air crew’s departure from the battlefield, instructed the aircraft
to engage this building and “destroy targets of all opportunity.” The target was actually a 6
Doctors Without Borders field hospital mistaken for a Taliban hideout.
The attack lasted for 29 minutes. By its conclusion, 30 Doctors Without Borders staff and
patients were dead and 37 injured. The death toll eventually climbed to 42 in what became one 7
of the deadliest attacks on civilians in the fifteen-year Afghan conflict. 8
Campbell, 2015.5
Rosenberg, Matthew. 2016. Pentagon details chain of errors in strike on afghan hospital. New York 6
Times. April 29, 2016. Campbell, 2015.7
BBC Newswire, 2016. Kunduz hospital bombing: 16 US forces ‘disciplined’.8
!2
Fisher
The Kunduz Strike and its Unexpected Connection to Defense Procurement
This event in Kunduz, at face value, has nothing in common with defense contracting.
The Pentagon assessed the Kunduz hospital bombing as the direct result of human error,
compounded by systems and procedural failures. Even so, investigations beyond the DoD 9
further probed system failures, and examinations of the Kunduz bombing soon criticized the
failings of Army intelligence tools developed by Beltway contractors.
Following the Pentagon’s press briefing on the Kunduz bombing, Representative Duncan
Hunter (CA-50) wrote a letter to Ashton Carter, the then-Secretary of Defense, criticizing the
DoD’s post-mortem analysis of the event. Hunter noted that the Distributed Common Ground
System (DCGS), an Army intelligence network developed by Lockheed Martin, was not
operational during the strike on Kunduz. Though a concrete link between DCGS downtime and 10
faulty intelligence in the Kunduz bombing has not been publicly established, Rep Hunter’s letter
touched upon preexisting skepticism of DCGS and its utility. A frequent critic of DCGS, Rep.
Hunter implored Carter to consider the limitations of the tool and the Army’s preference towards
this software. “Senior army leaders have gone to extraordinary lengths in recent years to deny
evidence of the failures of the DCGS program, and I am asking for your help to prevent them
from doing so following this tragic incident,” Hunter wrote. Hunter also advocated an 11
investigation over the Army’s reluctance to deploy an alternative intelligence analysis program—
Campbell, 2015.9
Tucker, Patrick. 2016. The war over soon-to-be-outdated army Intelligence Systems. Defense One. 10
July 6, 2016 BBC Newswire, 2016. Kunduz hospital bombing: 16 US forces ‘disciplined’.11
!3
Fisher
a software suite not developed by the likes of Lockheed Martin, Raytheon, or Northrup
Grunman, but developed by a Silicon Valley startup named Palantir Technologies. 12
Indeed, Lockheed’s Distributed Common Ground System (DCGS) was described by an
Army Test and Evaluation Command report as having “significant limitations” rendering the
service “not suitable and not survivable,” especially given its unreliability for special operations
troops. In contrast with criticism towards DCGS, special operations units had voiced a strong 13
preference for Palantir’s data analysis software; only two out of six special operations requests
for Palantir, however, were fulfilled, after members of Congress intervened. Following Rep. 14
Hunter’s investigation, a failed bombing in Afghanistan emerged as a potential manifestation of
contractor favoritism, demonstrating the consequences of procuring military tools from
contractors with close DoD connections, limited competition, and consequently thinned
incentives to produce the best tools possible.
The DoD’s relationship with large defense contracting firms stems from a history of
signing multi-billion dollar contracts for massive military expenditures, such as the design,
deployment, and maintenance of entire fleets of aircraft. Contracts of this scale could only be 15
fulfilled by large defense contractors, whose monetary and labor resources could fulfill detailed
military specifications and manufacture customized products. More recently, DoD acquisitions
have strayed beyond weapon and vehicle hardware to include data analysis software and other
novel technologies. Small startups can actually furnish the DoD’s demand for software and
lightweight hardware, in contrast to their inability to satisfy DoD munitions and vehicle
Mitchell, Ellen. 2016. How Silicon Valley's Palantir Wired Washington. Politico. August 14, 2016.12
Lyngaas, Sean. 2016. Army eyes DCGS reforms on Capitol Hill. FCW. June 1, 2016.13
Lyngaas, 2016.14
Department of Defense Instruction 5000.02. (2015).15
!4
Fisher
programs. Nonetheless, the Department has remained a stalwart customer of long-established
defense contractors, even for the development and purchase of newer categories of technology.
While young defense startups are known for their innovative products, the DoD’s multi-year
procurement process discourages small technology firms, who seek to avoid protracted
negotiations with distant payoffs. Agencies within the DoD have even been alleged of favoritism
towards traditional defense contractors. Cutting-edge products produced by the startup 16
community, despite their novel features and potential benefits, are often inaccessible for military
service people.
This distant relationship with small defense vendors comes at a time when the United
States grapples with insurgent warfare and cyber adversaries, whose strategies are dynamic and
embrace modern technologies. Recognizing the need to match adversaries and their emergent
technical expertise, the DoD has prioritized the inclusion of inventive—and even experimental—
software and hardware in military workflows. Incidents like the Kunduz hospital bombing
piqued interest in the potential of modern tools developed by young and rapidly growing defense
startups.
In 2015, then-Secretary of Defense Ashton Carter announced the creation of Defense
Innovation Unit Experimental (DIUx), an entity dedicated to procuring cutting-edge technologies
for agencies within the DoD. DIUx also attempts to establish relationships with small defense-
oriented companies and promote growth among them to stoke innovation in this market. A core
component of both goals above is the reformation of federal acquisition regulations, which have
long stood as a bureaucratic barrier to efficient defense procurement.
Tucker, Patrick. 2016. The war over soon-to-be-outdated army Intelligence Systems. July 6, 2016.16
!5
Fisher
DIUx, like a continuously developed product, follows “iterations” to test new strategies
for improved functionality and utility. Two years after the introduction of this program, this
thesis seeks to evaluate the efficacy of DIUx. More specifically, it assesses DIUx 2.0, the latest
iteration of an organization that claims to embrace the versioning and agility usually reserved for
small technology startups. DIUx 2.0, installed in May 2016, represented an overhaul in DIUx’s
leadership and strategy following Congressional and private industry criticism levied at the first
version of DIUx, DIUx 1.0; this thesis investigates whether DIUx 2.0 has made sufficient strides
towards DIUx’s mission to avoid additional substantial reforms. Specifically, this thesis pursues
the following questions about this DoD organization. Has DIUx 2.0 effectively facilitated
contracts between DoD agencies and innovative technology firms, and has DIUx offered utility
to defense technology startups? What are challenges facing DIUx’s current procurement strategy,
and are potential improvements for DIUx minor or substantial enough to mandate a version 2.1
or version 3.0, respectively?
Ultimately, this thesis finds that DIUx has effectively leveraged a specific procurement
vehicle, the Other Transaction Authority (OTA), in order to expedite contracts—but at the
possible expense of cultivating long-term relationships between defense customers and defense
technology startups. Addressing this tension, however, does not mandate a radical overhaul in
organizational leadership and strategy.
This thesis, by way of evaluating DIUx, also reinforces the importance of new
technology usage within the defense community, especially as newer technologies become
readily available on the public market and, by extension, available to adversaries like insurgents
—foes who previously did not have ready access to advanced tools.
!6
Fisher
Why the DoD’s Inclusion of Technology Startups Matters
American service members have expressed an interest in using technologies from
nontraditional defense vendors, especially as aggressors introduce the use of new off-the-shelf
software and hardware in the digital or physical battlefield. Additionally, research by the Small
Business Administration corroborates that small businesses generate new types of defense tools
that large Beltway contractors do not patent and develop as quickly.
Army personnel serving in Operation Enduring Freedom requested Palantir software,
which substantiated concerns over contractor favoritism in the Army’s preference for DCGS. A
2012 study performed by the Army Test and Evaluation Command indicated that the majority of
service people in this Operation preferred to use Palantir. 62% of respondents said that Palantir
was superior or greatly superior than these preexisting analysis tools, including DCGS. 17
Additionally, 91% of respondents said that Palantir increased their ability to support the
maneuver commander’s intelligence requirements versus previously deployed systems. This 18
last data point is especially revealing in the context of the Kunduz bombing, where DCGS
downtime hindered the intelligence available to the mission commander.
Secretary Carter also extolled the Pentagon’s need to keep pace with new technologies
not developed by the DoD or large defense contractors. Prior to his career in the Pentagon, Carter
had an academic career at Stanford University, an institution inextricably linked to the
technology startup community. Furthermore, before serving as the Secretary of Defense, Carter
Army Test and Evaluation Command Forward Operational Assessment Team XVIII. 2012. Forward 17
Operational Assessment: Palantir Operational Assessment Report. Fort Hood, Texas: Army Test and Evaluation Command. B-8. <https://www.wired.com/images_blogs/dangerroom/2012/08/1830_001.pdf>
Army Test and Evaluation Command Forward Operational Assessment Team XVIII. 2012. B-15.18
!7
Fisher
held the position of Under Secretary of Defense for Acquisition, Technology, and Logistics—a
role that assigned him responsibility over the DOD's procurement reform and innovation agenda.
With experience in both the Pentagon and Silicon Valley, Carter recognized the promise of novel
defensive technologies from vendors not well-known in the defense community. “Unlike when I
began my career and the best technology was coming from the Department of Defense, that’s all
changed. Now some of the best stuff comes from the outside, and we’ve sometimes been slow to
acknowledge that.” 19
The quick adoption of new commercial technologies by insurgent and terrorist groups has
accelerated the need for rapid response. Steve Blank, a professor of entrepreneurship at Stanford
and a former technician manager in the Air Force, highlighted this growing threat:
Today these potential adversaries are able to harness the power of social
networks, encryption, GPS, low-cost drones, 3D printers, simpler design and
manufacturing processes, agile and lean methodologies, ubiquitous Internet and
smartphones. Our once closely held expertise in people, processes and systems that we
once had has evolved to become commercial off-the-shelf technologies. U.S. agencies
that historically owned technology superiority and fielded cutting-edge technologies now
find that off-the-shelf solutions may be more advanced than the solutions they are
working on, or that adversaries can rapidly create asymmetric responses using these
readily available technologies […] All while we’re still writing a Request for a Proposal
from within the US Government procurement and acquisition channels. Today these
potential adversaries are able to harness the power of social networks, encryption, GPS,
Brill, Stephen. 2017. Donald Trump, Palantir, and the Crazy Battle to Clean Up a Multibillion-Dollar 19
Military Procurement Swamp. Fortune.
!8
Fisher
low-cost drones, 3D printers, simpler design and manufacturing processes, agile and lean
methodologies, ubiquitous Internet and smartphones. Our once closely held expertise in
people, processes and systems that we once had has evolved to become commercial off-
the-shelf technologies. 20
Reports of the Islamic State weaponizing commercial drones and using them to drop
explosive payloads on Iraqi Army outposts give credence to Blank’s concerns. Colonel Joseph 21
Felter, a research fellow at Stanford’s Center for International Security and Cooperation, also
found that over a three-year period in Afghanistan’s Helmand Valley, insurgent forces iterated on
new improvised explosive device (IED) deployments three times faster than American forces
could adopt new defensive strategies for IEDs. The gradual use of new IED detection and 22
mitigation technologies was able to address mounting IED attacks in this region, but this success
came after 4,899 IED casualties over three years. The expedited purchase and deployment of 23
new technologies can address emergent threats efficiently and minimize their damage.
Beyond defense community interest—among service people, leadership, and academics
alike—in new technologies from unconventional vendors, Small Business Administration (SBA)
research demonstrated that smaller companies are at the vanguard of developing and producing
new tools. A 2008 SBA study found that small firms, defined as those employing fewer than 500
Blank, Steve. 2016. Hacking for Defense @ Stanford – Making the World a Safer Place. <https://20
steveblank.com/2016/01/26/hacking-for-defense-stanford/> Warrick, Joby. 2017. Use of weaponized drones by ISIS spurs terrorism fears. The Washington Post. 21
February 21, 2017. Felter, Col Joseph. 2017. Solving National Security Issues with the Lean Launchpad. Presentation at 22
Hacking for Defense and Hacking for Diplomacy Educatiors & Sponsors Course, Georgetown University, Washington, DC
Felter, 2017.23
!9
Fisher
people, are much more likely to develop emerging technologies than large firms. Though small 24
firms account for only 8 percent of patents granted, they account for 24 percent of the patents in
the top 100 emerging clusters of technologies. This means that small firms produce three times 25
as many patents as one would expect in this special patent set. Put another way, approximately
one in 31 small firm patents are contained in the top emerging clusters, compared with one in
117 large firm patents. 26
To fold cutting-edge technologies into defense workflows, it is imperative for the DoD to
recognize the prevalence of small firms in cutting-edge technology sectors. Interactions with
startup companies are crucial for evaluating and sourcing high-technology defense tools. In its
mission to better equip the DoD and procure new technologies from promising young technology
firms, DIUx endeavors to establish a closer working relationship between the Pentagon and these
startups.
Methodology and Overview
This thesis will examine DIUx along three axes—(1) the utility of DIUx’s role in an
already crowded defense innovation ecosystem, (2) the efficacy of DIUx’s first round of
organizational reform in 2016, and (3) actions taken to identify and address emerging issues from
DIUx 2.0 and its first quarter of contracts.
To perform the first assessment, this thesis’s second chapter examines sources of
potential redundancy between DIUx and preexisting defense innovation programs. This chapter
Breitzman, Anthony, and Diana Hicks. 2008. An Analysis of Small Business Patents by Industry and 24
Firm Size. Haddonfield, NJ: Small Business Administration. i. Breitzman, Anthony, and Diana Hicks. 2008. v.25
! Breitzman, Anthony, and Diana Hicks. 2008. v.26
!10
Fisher
also serves as a literature review on already extant innovation efforts and their ability to yield
new technologies for the DoD. Through this analysis, this chapter finds that DIUx’s mission to
streamline procurement from defense startups addresses challenges in turning over new
technologies from DARPA and In-Q-Tel to DoD customers.
To examine the second axis, this thesis’s third chapter evaluates the organizational shift
from DIUx 1.0 to DIUx 2.0 and the utility of this change towards meeting DIUx’s twofold
mission: streamlining federal acquisition regulations and expediting new technology purchases.
To examine this success rate, this research takes into account the quantity and speed of DIUx-
managed contracts before and after this reform, and compare these procurement timeframes to
those of traditional contracting efforts. This thesis finds that DIUx’s heavy utilization of OTAs is
responsible for an uptick in DIUx contract count and speed.
Finally, this thesis’s fourth chapter seeks to identify continuing or novel barriers between
DIUx’s current state and its organizational mission. To locate these concerns, this thesis draws
upon three case studies, rooted in observations of DIUx meetings and interviews with personnel
from DIUx and the private companies they contract. This qualitative data exposes emergent
concerns about DIUx’s reliability on OTAs, whose short-term contracts are not immediately
conductive to long-term business relationships between the DoD and defense technology
startups.
This approach is intended to create a holistic analysis of DIUx’s role in the defense
community, its evolution as an organization, and the continuing challenges DIUx and young
defense vendors face—all pressing issues as the need for novel defense technology mounts.
!11
Fisher
!12
Fisher
Chapter II | Surveying the Defense Innovation Landscape Before DIUx
To contextualize DIUx’s role in an already populated defense innovation ecosystem, an
analysis of preexisting defense innovation programs—in addition to their goals and shortcomings
—is prudent. As such, an overview of research pertaining to older defense innovation programs
can inform a symbiotic role for DIUx among its predecessors. This overview can also define
optimal outcomes for this new DoD program, as imparted by the previous failures and successes
of other defense innovation efforts.
Academic research on defense innovation focuses on four types of actors: equity
investments through Government Venture Capital (GVC) programs, work programs between
government agencies and private companies, the ability for federal acquisition regulations (FAR)
to encourage the use of novel products, and rapid acquisition task forces that utilize FAR
discretionary authorities for quick procurement. GVC outfits make equity investments in small
companies, typically receiving an observer board seat to oversee and provide advice to their new
portfolio business. Work programs, much like the GVC approach, provide funding to 27
companies and R&D programs; this type of investment, however, is conditioned upon the
funder’s R&D agenda. Efforts to streamline federal contracting, unlike GVC and work 28
programs, include Congressional efforts to amend FAR and increase their clarity and efficiency. 29
Brander, James, Qianqian Du, and Thomas Hellman. 2010. The effects of government-sponsored 27
venture capital: International evidence. National Bureau of Economic Research (November). 572. Yannuzzi, Rick. 2000. In-Q-tel: A new partnership between the CIA and the private sector. Defense 28
Intelligence Journal 25 (26). Searle, David, David Christensen, and Caisse Vickery. 1999. The impact of the packard commission's 29
recommendations on reducing cost overruns on defense acquisition contracts. Acquisition Review Quarterly (Summer). 251.
!13
Fisher
The success of FAR depends on federal agencies’ implementations of these regulations,
particularly through rapid acquisition task forces. 30
Given that newer organizations like DIUx have not yet received academic scrutiny, the
majority of research specific to defense and intelligence innovation programs examines older
agencies such as DARPA, in addition to publicly funded non-profits like In-Q-Tel. GVCs are
relatively popular subject matter in business and management academia, while federal auditing
and government program analyses focus on work programs and FAR reforms. Though DIUx is
not explicitly the subject matter in this research, academic critiques of established programs can
enumerate successes and challenges in defense and intelligence innovation programs, and
identify continuing issues for DIUx to address.
A Brief Overview of In-Q-Tel and DARPA
In-Q-Tel and DARPA are among the US Government’s most prominent defense and
intelligence innovation organizations. In-Q-Tel, a non-profit established by the CIA in 1999, is a
GVC fund, but also performs work program investments. The In-Q-Tel Interface Center (QIC) is
an organization within the CIA that oversees In-Q-Tel’s investments; it primarily functions as a
junction point between the GVC’s unclassified efforts and classified intelligence community
work. Given its ability to generate its own value through portfolio investments, In-Q-Tel seeks 31
Searle et al., The impact of the packard commission's recommendations on reducing cost overruns on 30
defense acquisition contracts. 251. Yannuzzi, In-Q-tel: A new partnership between the CIA and the private sector.31
!14
Fisher
to reach a financial state in which it can fund its own operational costs and front its own
investments without additional federal funding. Its assets in FY 2014 were $340 million. 32 33
DARPA, established in 1958 to accelerate the research and development of US military
technologies in the Cold War, is an agency within the DoD that continues to spearhead the
Department’s development and deployment of emerging technologies. Through agency
announcements, DARPA publicly solicits and assists with the development of new technologies
for DoD needs. DARPA does not make equity investments like a GVC, but engages in work
program investments. R&D groups that successfully win DARPA research awards can receive
grants typically between 220,000 to ten million dollars in size. Its FY 2016 budget was $2.87 34
billion. 35
Defense GVC Programs
The success of GVC programs is often assessed in terms of a portfolio company’s
revenue growth, which increases that company’s value and generates a return on investment
(ROI) for its government investor. Defense GVCs, however, also prioritize the adoption of
portfolio companies’ products among federal agencies. In order to encourage the production of
higher-quality, commercially viable goods, GVC programs also attempt to stoke economic
competition among small technology companies. 36
Yannuzzi, In-Q-tel: A new partnership between the CIA and the private sector.32
In-Q-tel form 990. 2013.33
Wardle, Caroline. 1999. Obtaining federal funding. Arlington, VA: National Science Foundation. 29.34
Senate Armed Services Committee, Subcommittee on Emerging Threats and Capabilities, Strategy and 35
Implementation of the Department of Defense's Technology Offsets Initiative in Review of the Defense Authorization Request for Fiscal Year 2017 and the Future Years Defense Program, 114th sess., 2016.
Wonglimpiyarat, Jarunee. 2006. The dynamic economic engine at silicon valley and US government 36
programmes in financing innovations Technovation (26). 1082.
!15
Fisher
I. Improving Returns on Investment and Returns on Technology
Management scholar Josh Lerner and venture capitalist Felda Hardymon wrote that
national security-oriented GVCs attempt to maximize a “return on technology,” defined by In-Q-
Tel as the ratio of In-Q-Tel portfolio company products assimilated by intelligence community
clients. Indeed, In-Q-Tel seeks to furnish the CIA with cutting-edge technology from a 37
competitive marketplace offering suitable and readily deployable off-the-shelf products. In
pursuit of this goal, In-Q-Tel invests in technology companies to support the development of
commercial, readily usable technologies with national security uses, thereby discouraging
bespoke solutions that have previously resulted in protracted contracts. Lerner and Hardymon
described the character of In-Q-Tel’s operations as less like a conventional venture capital fund
and more like a corporate strategic venture fund, given In-Q-Tel’s intent to cannibalize the
technology it invests in. Acting like a corporate strategic venture fund obliges In-Q-Tel to build 38
relationships with the emerging companies it invests in, and assist in the company’s design of a
commercially viable product that would also fulfill CIA needs. Additionally, the authors stated 39
that investments in multiple companies in the same product market stimulate these firms to
compete and further develop their wares, from which the CIA can purchase the best product.
Lerner and Hardymon described this tactic through the words of former In-Q-Tel CEO Gilman
Louie:
Lerner, Josh, and Felda Hardymon. 2005. In-Q-tel: Harvard business school case 9-804-146. 1.37
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 1.38
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 9.39
!16
Fisher
Basically, this model lets the Agency “taste before it buys” and uses other
people’s money to help make the goods. In our model, there wasn’t enough money to
solve all of the Agency’s problems under a direct procurement approach—$40 million
wasn’t going to get you that far. The strategy was to leverage the outside corporate
strategic and conventional venture funds by joining up in a traditional syndicate model.
As a result, the CIA didn’t have to pick a single winner or loser. Now they could
bet on three or four companies in a given product space without having to go through the
process of a single procurement, which was good because it was very hard to pick
winners in new technologies, and the cycle times were very short, particularly in the IT
space.
We bet in five to six spaces, 20 to 25 times per year. The best companies will rise
to the top, and we can quickly vet out the ones that aren’t so good, because the ones that
aren’t as good aren’t going to get financial support. And even if we choose wrong, we
will know what our competition is, so the Agency will be able to make procurement
decisions with better information. 40
Additionally, Lerner and Hardymon featured the In-Q-Tel Interface Center as a significant actor
in the successful adoption of products from In-Q-Tel portfolio companies. The QIC provides In-
Q-Tel with a list of unclassified technology needs on an annual basis, and this list—called the
Problem Set—defines the sectors in which the organization would invest and seek to procure
technology. Lerner and Hardymon noted that the QIC Problem Set ideally serves as “a cultural
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 5.40
!17
Fisher
convergence point for In-Q-Tel employees regardless of geographic location and professional
skill set,” and described the Center as a critical role-player in transferring In-Q-Tel’s solutions to
Agency customers. 41
Andrew S. Mara, a former AAAS Science and Technology Policy Fellow, found that
framing In-Q-Tel as a corporate strategic venture fund is still insufficient to maximize returns on
technology. He asserted that equity investments and overseer board positions alone do not
promote the adoption of new, defensive technologies among federal agencies. QIC’s focus on 42
developing a Problem Set for In-Q-Tel, though useful for targeting In-Q-Tel’s investments to
ensure that they serve a CIA function, also failed to ensure that In-Q-Tel portfolio company
products actually receive a contract and get assimilated into CIA workflows. In a survey of 39 43
In-Q-Tel portfolio companies, only 26.3 percent of companies thought that they were highly or
extremely likely to sell to the government prior to their In-Q-Tel interactions. This number
substantially increased to 78.9 percent after In-Q-Tel interactions. Mara found, however, that a 44
maximum of 25 percent of companies in the In-Q-Tel portfolio go on to win government
contracts after investment, and argues that this return on technology can be increased with
heightened efforts to encourage government agencies to use off-the-shelf solutions.
Given that interactions with In-Q-Tel were highly effective in encouraging small private
companies to collaborate with and sell to the government, Mara encouraged GVC programs like
In-Q-Tel to share information about portfolio companies’ products with other agencies and
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 7.41
Mara, Andrew. 2011. Maximizing the returns of government venture capital programs. DEFENSE 42
HORIZONS National Defense University (71) (January). 5. Mara, Maximizing the returns of government venture capital programs. 5.43
Mara, Maximizing the returns of government venture capital programs. 5.44
!18
Fisher
specifically recommended the creation of a searchable database containing contact information,
detailed technology descriptions, and product field test results for each company in their deal
flow. According to Mara, GVC programs also need to develop methods to reliably produce 45
“technology champions” in government agencies who advocate the adoption of readily available,
commercial products. He suggested that these “champions could be personnel borrowed from
R&D efforts [within government] who […] shepherd in new technologies.” Mara also argued 46
that project managers could be directly involved in portfolio company selection linking
successful technology integration—and not just successful product development—to the project
manager’s own goals and priorities. 47
II. Stimulating Competition and Innovation in Defense Products
The economist Michael Porter, who specializes in management analysis, evaluates GVCs
in terms of clusters of geographically proximate businesses. He claimed that clusters best
represent the influence of competition and government investments in a specific industry,
arguing that this model captures important linkages, complementarities, and spillovers in terms
of technology, skills, information, marketing, and customer needs that cut across firms and
industries. Porter represented the effect of clustering using the following diamond model, 48
whose corners represent conditions that catalyze product innovation and production, and whose
links are factors contributing to ideal conditions.
Mara, Maximizing the returns of government venture capital programs. 6.45
Mara, Maximizing the returns of government venture capital programs. 8.46
Mara, Maximizing the returns of government venture capital programs. 8.47
Porter, Michael. 2000. Location, competition, and economic development: Local clusters in a global 48
economy. Economic Development Quarterly 14 (1) (February). 16-18.
!19
Fisher
This model includes the success factors identified by Lerner, Hardymon, and Mara. One
link, “vigorous competition among locally based rivals,” is congruent with GVC ambitions to
use investments to foster competition among companies in an emerging market. Similarly, the
link “unusual local demand in specialized segments” aligns with Mara’s suggestion that GVC
programs encourage federal agencies to specifically adopt portfolio companies’ products as
commercial solutions to agency needs. The link “customers’ needs that anticipate those
elsewhere” also describes Mara’s advocacy of technology champions, whose advocacy of off-
the-shelf products ideally extends beyond the intelligence community to engage a wider audience
of federal agencies.
Of the other success factors outlined in the diamond model, Porter recognized
government policies towards intellectual property and foreign investments as substantial factors
in evaluating GVC programs and their ability to support new companies:
The investment climate and policies toward competition set the context. Things
such as the macroeconomic and political stability, the tax system, labor market policies
!20
Fisher
affecting the incentives for workforce development, and intellectual property rules and
their enforcement contribute to the willingness of companies to invest in upgrading
capital equipment, skills, and technology. Antitrust policy; government ownership and
licensing rules; and policies toward trade, foreign investment, and corruption have a vital
role in setting the intensity of local rivalry. 49
Porter’s emphasis on GVC programs and their support of companies’ intellectual property
(IP) touches upon a concern among new companies that, if they were to accept investments from
organizations linked to the defense or intelligence community, their IP could be used in a
classified national security operation, revoking their ability to use their IP in public. The 50
classification of IP would forbid the public sale of a product derivative of this IP, and limit a
company’s market expansion opportunities. In-Q-Tel, however, offers its portfolio companies 51
IP protections, both legal and in practical: this GVC’s mission advocates the cultivation of
commercially viable products, not products whose IP and use is restricted to the CIA. 52
Additionally, a start-up may believe that a relationship with an organization like In-Q-Tel, and its
explicit relationship to the US Intelligence Community, could complicate the possibility of
foreign investment in said company. In-Q-Tel, nonetheless, is not prohibited from investing in
non-US companies and states it will consider such investments as appropriate based on the fit
with the needs and priorities of IQT and the IC. This ability to invest in a foreign company 53
Porter, Location, competition, and economic development: Local clusters in a global economy. 21.49
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 12.50
Lerner and Hardymon, In-Q-tel: Harvard business school case 9-804-146. 12.51
In-Q-Tel. Submit a business plan FAQ. in In-Q-Tel [database online]. Arlington, VA, [cited 11/22 52
2016]. Available from https://www.iqt.org/business-plan-submission-guidelines/submit-a-business-plan-faq/.
In-Q-Tel, Submit a business plan FAQ.53
!21
Fisher
suggests that In-Q-Tel is not principally opposed to an investment position alongside foreign
investors, potentially assuaging concerns that In-Q-Tel investments may block an entire class of
other investment opportunities.
Broadly summarizing the academic arguments above, the success of a GVC-based
approach to developing and deploying new defensive technologies hinges on two principal
outcomes. First, an effective defense GVC generates a return on technology, such that portfolio
company products prove useful to defense and intelligence community clients, and are absorbed
into federal agency processes and practices. A successful defense GVC also invests in companies
intending to create a commercially viable product refined by economic competition. Though
there is no academic consensus on which factors most significantly influence the two outcomes
above, these factors can include the level of autonomy given to GVC portfolio companies,
awareness of GVC efforts, the presence of “technology champions” to promote the use of GVC
portfolio companies’ tools, and the competition a GVC seeks to cultivate in product sectors of
national security interest.
Defense Work Programs
Defense work programs focus on the cultivation of R&D with the potential to serve
national security needs. In-Q-Tel has a work program that invests in companies whose products
are not immediately deployable, but can better suit IC needs through additional In-Q-Tel input
and advising. In-Q-Tel’s narrow scope for product development differs from the broad approach
espoused by DARPA, the DoD’s most prominent R&D support outfit. Instead of identifying
specific products for work programs, DARPA defines and supports entire sectors of technical
!22
Fisher
research that contribute to national security needs. These sectors include a variety of research 54
areas including, but not limited to, autonomous vehicles, improved prosthetics, and drones.
Analysis of In-Q-Tel and DARPA work programs identifies two ideal outcomes: the stimulation
of advanced defense R&D, and the eventual federal procurement of products developed through
work programs. These two goals are sometimes are at odds, especially as cutting-edge R&D
becomes difficult to integrate into defense programs.
I. Cultivating R&D to Advance Defense Technologies
Unlike a GVC firm like In-Q-Tel, DARPA prioritizes the support of R&D over returns on
investment. DARPA drafts its research agenda independent of other DoD agencies, assigns
program managers to particular fields of research, then posts announcements to solicit proposals
for relevant R&D projects. Successful proposals receive DARPA grants, typically ranging from 55
six to seven figures in value. Though these grants are in the form of contracts, they are not for 56
product acquisition, but for funding research that fulfills DARPA interests.
Van Atta et al. at the Institute for Defense Analysis argue that DARPA’s autonomy is
conducive to innovation. In an assessment of the program, they describe DARPA’s approach to
R&D projects as high-risk with a high payoff, and laud DARPA’s independence from military
specifications. “DARPA program managers are encouraged to challenge existing approaches to
warfighting and to seek results rather than just explore ideas,” Van Atta writes, finding that a
Defense Advanced Research Projects Agency. About DARPA. Arlington, VA, [cited December 12 54
2017]. Available from http://www.darpa.mil/about-us/about-darpa. Defense Advanced Projects Research Agency. Contract management. Arlington, VA, [cited December 55
12 2017]. Available from http://www.darpa.mil/work-with-us/contract-management. Wardle, Obtaining federal funding. 29.56
!23
Fisher
crucial pillar of DARPA’s work is to conceive a future beyond existing DoD tools and
practices. He describes DARPA’s R&D success in terms of the overall innovation it spawns, 57
rather than the defense community’s adoption of DARPA-backed products: “DARPA’s steady,
forward-looking promotion of critical technologies—before their national security significance
becomes clear—has supported US dominance of entirely new industries, e.g., microelectronics,
advanced computing, networking, and other information technology industries.” 58
Vehse et al., who received grants from DARPA in their capacity as researchers at AT&T
Bell Laboratories, stated that DARPA’s independent R&D agenda allowed for broad research
objectives to which multiple grant winners could concurrently contribute. These independent 59
R&D groups can competitively research similar subsets of a research topic or perform
independent R&D complementary to each other’s efforts. DARPA cultivates overarching 60
research objectives to increase the number of teams that can perform R&D, attempting to reap as
many research results possible.
Despite the liberties DARPA’s sweeping research objectives granted researchers, Vehse et
al. claim that broad research projects detracted from the ability for DARPA managers to
effectively support individual R&D teams. “Each project task had to be defined at the highest
level, using information found in the Statement of Work and other contract documents,” Vehse
describes, “and at the ‘working level,’ resource commitments and milestone choices had to be
Van Atta, Richard, Michael Lippitz, Jasper Lupo, Rob Mahoney, and Jack Nunn. 2003. Transformation 57
and transition: DARPA’s role in fostering an emerging revolution in military affairs. Alexandria, VA: Institute for Defense Analyses. S-1.
Van Atta et al., Transformation and transition: DARPA’s role in fostering an emerging revolution in 58
military affairs. S-1. Vehse, Robert, Stephen Nygren, and Duane Butherus. 1991. Managing an R&D contract with the 59
government. AT&T Technical Journal (March/April). 87. Vehse et al., Managing an R&D contract with the government. 8760
!24
Fisher
consistent with the Work Breakdown Structure and the master implementation schedule for the
entire project.” In Vehse’s experience, R&D groups were left to interpret contract conditions 61
and research requirements for overarching DARPA research initiatives, then apply those findings
to their highly specific R&D projects. To closely collaborate with DARPA personnel and discuss
research in specific terms, R&D teams had to become self-starters about opening up and
maintaining constant lines of communication with relevant DARPA managers. 62
In the agency’s attempt to maximize independent and risky research with possibly high
payoffs, DARPA focused on high-level research announcements, maintaining loose ties with
R&D teams like Vehse’s. As a work investment program, DARPA did not initially provide hands-
on support or feedback beyond their investment in a group’s research proposal, allowing
researchers to operate as autonomously as DARPA operates within the DoD.
II. Procuring Advanced Technologies from R&D Supported by Work Programs
The Government Accountability Office, in recent reports on DARPA performance, has
criticized the agency’s emphasis on high-risk, independently pursued research. A 2015 GAO
report stated that “[DARPA] programs generally seek to prove the art of ‘what is possible’ rather
than refining, producing, and delivering tactical equipment to warfighters.” The report 63
continues,
Vehse et al., Managing an R&D contract with the government. 9161
Vehse et al., Managing an R&D contract with the government. 9362
Government Accountability Office. 2015. DEFENSE ADVANCED RESEARCH PROJECTS 63
AGENCY: Key factors drive transition of technologies, but better training and data dissemination can increase success. Washington, DC: Government Accountability Office, GAO-16-5. 18.
!25
Fisher
[…] the introduction of DARPA’s radically innovative technologies can disrupt
the status quo for military programs, budgets, and warfighting doctrine, which can drive
cultural opposition within the military services. DARPA officials stated that the agency’s
research sometimes leads to the identification of technologies and capabilities that
military service officials do not initially want or think their services will need, although
these technologies can eventually provide important military capabilities. 64
Advocating a renewed focus on the transition of DARPA R&D to practical DoD use, the
GAO recommended that DARPA’s leadership and program managers prioritize the deployment
of DARPA-funded tools. The utilization of DARPA-funded R&D was a responsibility previously
assigned to DoD management above DARPA, including the Director of Defense Research and
Engineering, the Under Secretary of Defense for Acquisition, Technology and Logistics, and the
Secretary of Defense. Because DARPA program managers are well-versed in DARPA’s 65
research portfolio, the GAO argued that these staffers would better interface DARPA and defense
community members. They can convey DoD needs in the DARPA research agenda, and link 66
DARPA-funded researchers to possible defense community customers. Moving towards a
DARPA where program managers facilitate the transition of R&D, the agency can include
military liaisons from the Air Force, Army, Navy, and Marine Corps in the drafting of DARPA’s
research agenda, so that defense community priorities are better reflected in DARPA’s research
Government Accountability Office, Key factors drive transition of technologies, but better training and 64
data dissemination can increase success. 19. Van Atta et al., Transformation and transition: DARPA’s role in fostering an emerging revolution in 65
military affairs. S-12. Government Accountability Office, Key factors drive transition of technologies, but better training and 66
data dissemination can increase success. 22.
!26
Fisher
announcements. Military liaisons can also link DARPA program managers to potential 67
transition partners or other stakeholders in the military services. The GAO report also 68
recommends that DARPA managers should pursue the formal training and certification needed to
perform government contracting, so they can advise DARPA grant winners throughout the
procurement process. Despite efforts to integrate DARPA-funded R&D into DoD workflows, 69
fewer than 1% of DARPA-funded products are transitioned to Pentagon usage. 70
As DARPA moves towards its sixth decade of operation, the agency has the challenge of
balancing high-risk, high-reward research goals against a growing pressure for R&D that can
serve the DoD’s immediate interests. Defense work programs, when established under a broad
mandate to advance R&D, have historically cultivated entirely new technology sectors.
Sweeping research objectives, nonetheless, are not optimized for the rapid deployment of new
products to the defense or intelligence community. Without clear procurement strategies or
support in place, DARPA possibly leaves a crucial junction of the defense innovation pipeline—
the contracting of innovative tools—without service.
Federal Acquisition Regulations
Federal Acquisition Regulations, which define procurement vehicles for the federal
government, underpin the actions the defense and intelligence communities can take to contract
Government Accountability Office, Key factors drive transition of technologies, but better training and 67
data dissemination can increase success. 22. Government Accountability Office, Key factors drive transition of technologies, but better training and 68
data dissemination can increase success. 22. Government Accountability Office, Key factors drive transition of technologies, but better training and 69
data dissemination can increase success. 21. Tseng, Brandon. Interview by author. May 2, 2017.70
!27
Fisher
products. FARs are not written by federal agencies inasmuch they are established and updated by
Congress. Recent FARs have sought to curb runaway contract costs and timelines, while
advocating procurement relationships with smaller, less-established defense technology
companies. Broad metrics for successful contracts, as cumulatively defined by fifty pieces of
FAR legislation and FAR initiatives, involve the minimization of contract breaches, the
promotion of economic competition between contractors, and support towards small businesses
attempting to gain federal customers. These FARs grant discretionary authority for the DoD to 71
implement acquisition practices that Congress endorses. Ultimately, the discretionary use of FAR
to streamline procurement is performed by rapid acquisition task forces within DoD agencies.
I. Discouraging Contract Breaches
Certain contracts—as such as the $400 billion F-35 acquisition program—have become
emblematic of delayed schedules and inflated costs in defense procurement. The Nunn-McCurdy
Provision of 1983 sought to curb runaway contract schedules and prices by forcing the
cancellation of a program once it reached certain cost growth thresholds. This provision,
however, holds little leverage: Nunn-McCurdy has never been used to cancel a contract, and
instead serves as a tripwire for Congressional inquiry into that transaction’s necessity. Even 72
when a Congressional inquiry is prompted, the DoD can submit notifications and certifications of
a contract’s importance to indefinitely delay Congress from cutting funds. The House and Senate
Armed Services Committees can also prevent Congress from cutting funds in the wake of a
See Appendix A for a list of all Acquisition Reform initiatives and FARs passed in the United States 71
Congress. Schwartz, Moshe, and Charles O'Connor. 2016. The Nunn-McCurdy Act: Background, analysis, and 72
issues for congress. Washington, DC: Congressional Research Service.
!28
Fisher
Nunn-McCurdy breach. Jonathan Ritschel of the Air Force Institute of Technology argues that if
there is little or no penalty to a Nunn–McCurdy breach, then bureaucratic behavior will remain
mostly unchanged. Given the low stakes of such a breach, Ritschel’s hypothesis about the weak
deterrence of this provision appears to hold. FARs alone fail to curtail runaway procurement 73
prices, and present another success factor for defense innovation programs: an intrinsic prudence
towards spending and contract management.
Daniel Czelnik and Philip Rogers, both analysts in the Office of the Undersecretary for
Defense for Acquisition and Technology, suggested that variable funds for contingency
operations can contribute to cost overruns. Indeed, their models revealed that war positively
correlates with overruns that could constitute a Nunn-McCurdy breach. During periods of 74
unpredictability in defense community needs, procurement innovation initiatives can attempt to
reduce the number of contract breaches that occur, using an agile and responsive procurement
process in tune with the DoD’s rapidly evolving challenges.
Valerie Grasso of the Congressional Research Service argued that commercial products
can frequently fulfill defense community demands. Commercial solutions, she found, cost less 75
and have comparable quality to bespoke products built to DoD military specifications. Like
Czelnik and Rogers, Grasso asserted that the DoD’s procurement specifications do not keep
contractors abreast of changing requirements. High-technology commercial products can evolve
Ritschel, Jonathan. 2012. Efficacy of US legislation in military acquisition programmes: Nunn–73
McCurdy act unveiled. Economic Papers 31 (4) (December). 496. Cooper, Michael. 2002. Department of defense implementation of acqusition reform initiatives. 74
Master's Thesis., Naval Postgraduate School. 59-65. Grasso, Valerie. 2000. Defense acquisition reform: Status and current issues. Washington, DC: 75
Congressional Research Service.
!29
Fisher
so quickly that Grasso contended that the DOD’s “milspec" system cannot keep pace with these
new and potentially valuable advances. 76
David Christensen et al. at the Air Force Institute of Technology found that development
contracts are more risky than production contracts, and appear to be more sensitive to policies
affecting cost performance. Production contracts are related to the manufacturing of preexisting 77
products as opposed to the development of new ones according to military specifications.
Christensen’s caution towards protracted development contracts complements Grasso’s advocacy
for off-the-shelf solutions.
A rapid and responsive procurement process is a widely-recognized objective for defense
contracting. While long-term contracts are criticized for their tendency to overrun without
powerful mechanisms to discourage these breaches, short-term contracts for commercial
solutions apply to a class of products conductive to smaller contract prices and timeframes.
Intelligence and defense innovation programs like In-Q-Tel and DIUx similarly advocate for the
rapid acquisition of off-the-shelf products, and work towards actualizing this goal when
Congressional mandate alone cannot.
II. Stimulating Competition Among Contractors
The Federal Acquisition Streamlining Act of 1994, and its following iterations, sought to
empower the acquisition workforce to make decisions about how to manage their programs.
FASA followed the Packard Commission, a federal government commission commissioned by
Grasso, Defense acquisition reform: Status and current issues.76
Searle et al., The impact of the packard commission's recommendations on reducing cost overruns on 77
defense acquisition contracts. 257.
!30
Fisher
President Reagan to evaluate management in the DoD. This Commission recommended
streamlining the acquisition process by increasing testing and prototyping of possible solutions
and modeling the DoD after competitive firms with a history of efficient acquisitions. FASA 78
implemented some of the Commission’s findings; its call for better-trained and more autonomous
procurement personnel constituted a shift away from more restrictive and centralized
procurement practices of the past. Additionally, FASA advocated the use of performance based 79
contract payments, and urged procurers to consider commercially available products. 80
Complemented by the Competition in Contracting Act of 1984, procurement officials’
stimulation of economic competition between contractors emerged as another criterion of
effective FAR. Competition incentivizes defense technology companies to produce better 81
products faster, at a lower cost, and in larger quantities—a boon for defense community members
seeking high-quality tools without contracting markups or overruns.
To encourage competition between contracting companies, the Indefinite Delivery
Contract (IDC) became a popular item in the FAR toolbox. An IDC is a vehicle that has been
awarded to one or more vendors to facilitate the delivery of supply and service orders. This type
of contract does not procure or specify a firm quantity of services (other than a minimum or
maximum quantity) and provides for issuance of orders for the performance of tasks during the
period of the contract. IDCs give the federal government the ability to expand the list of 82
Searle et al. The impact of the packard commission's recommendations on reducing cost overruns on 78
defense acquisition contracts. Cooper, Department of defense implementation of acquisition reform initiatives.79
Cooper, Department of defense implementation of acquisition reform initiatives. 80
Gansler, James, William Lucyshyn, and Michael Arendt. 2009. Competition in defense acquisitions. 81
College Park, MD: University of Maryland Center for Public Policy and Private Enterprise. Federal Procurement Data System. Indefinite delivery contract. Washington, DC, [cited November 29 82
2017]. Available from https://www.fpds.gov/help/Indefinite_Delivery_Contract.htm.
!31
Fisher
required services expected of a contract, such that dynamic federal agency needs can be reflected
in procurement agreements. Because this contracting vehicle can award multiple competitors at
once, the IDC can stoke competition between these companies to provide the best product in
order to actually secure the payments they have booked in a contract.
IDCs facilitate the economic competition that FARs encourage, though they particularly
disadvantage smaller companies who may not have sufficient resources to respond to shifting
defense community needs versus wealthier, better-established Beltway contractors. Supporting
start-up businesses in the realm of defensive technologies is crucial to welcoming the newest
technologies into defense and intelligence workflows, and presents another domain for defense
innovation initiatives to address.
III. Supporting Small Businesses to Capture New Technologies
Following the adoption of IDCs, smaller American contractors struggled to compete with
larger contracting firms experienced with adjusting commercial products for government needs.
Their inclusion in the contracting fold is Congressionally mandated: the 1997 Small Business
Act mandates that 23% of procurement dollars go to what qualifies as a small business. The DoD
was specifically instructed by Congress to have contracting officers solicit offers from any size
of contractor that offered the required services under a contract. No conclusive link, however,
exists between the mere act of including smaller companies in contract considerations and the
actual awarding of a contract to a smaller firm. The presence of smaller companies among 83
Clark, Major, and Chad Moutray. 2004. The future of small businesses in the U.S. federal government 83
marketplace. Journal of Public Procurement 4 (3). 452.
!32
Fisher
awarded contracts—and their propensity to introduce cutting-edge technologies—remains a
crucial goal among defense community innovation efforts.
Karen Thornton, a former procurement attorney with the US Army Corps of Engineers,
examined components of IDCs that detriment small companies seeking federal contracts. She
found that federal agencies may poorly estimate quantities and insufficiently define product
requirements in initial contracts, which prompts additional task orders in an agreement—to the
detriment of smaller companies. Startups have less capital for rapidly customizing products in 84
response to updates in DoD specifications; as a result, they may not be able to extend contracts
beyond the terms of initial agreement. Thornton argued, “successful multiple award contracting
relies in large part on contractors continuing to submit their best proposals throughout the
duration of the contract. This can be an expensive proposition for contractors. Because agencies
often over- or under-assess their estimates, the built-in maximum and minimum quantities clause
is intended to protect a contractor from being forced to perform at a rate that exceeds its capacity
or causes underperformance.” Unpredictable contract terms and estimates can discourage 85
smaller companies from even selling their wares to federal customers. 86
Another issue Thornton identified in FAR is FASA’s preclusion of protests against task
order awards—a view she shared with Major Clark, former Assistant Chief Counsel in the US
Small Business Administration. Clark asserted that “the ability of a small business to protest 87
Thornton, Karen. 2002. Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite 84
delivery contract. Public Contract Law Journal 31 (3) (Spring). 410. Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 85
contract. 410. Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 86
contract. 411. Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 87
contract. 407.
!33
Fisher
and win in a conflict in which an agency has decided to incorporate functions into an existing
multiple-award, indefinite-delivery contract is extremely slim.” Because small businesses are in 88
a position where federal consumers exert control over contract terms, they may cast departments
like the DoD as recalcitrant customers. Thornton wrote, “with limited access to protest forums,
offers might be less willing to take the risk associated with competing. More than an overspend
mechanism, the bid protest system is the most efficient method of policing the contract formation
process.” Without robust bid protest systems in place, smaller contractors navigate the 89
procurement space while lacking a metaphoric seat at the contract negotiating table.
In a landscape where small companies, if seeking business from the federal government,
must compete against established beltway contractors, Thornton and Clark each argued that
small companies would benefit from an explicit “fair opportunity” process when federal
procurement groups award and administer award task and delivery order contracts. They also 90
agreed that federal agencies should attempt to furnish contracts that accurately portray their
customer needs in the present and near future, therefore minimizing the chance of unexpected
task orders beyond the original scope of the contract. The federal government’s procurement
officers, Clark argued, should heighten their ability to craft tightly focused contracts with clear
expectations—the form of contract most accessible to small business bidders. To achieve this
goal, the acquisition workforce must be trained about the full span of possible contractual
vehicles and their unique appeals to smaller companies. Thornton envisioned the ideal
Clark et al., The future of small businesses in the U.S. federal government marketplace. 459.88
Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 89
contract. 408. Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 90
contract. 413.
!34
Fisher
contracting officer workforce as one that promotes competition among all contractors as opposed
to the large contracting firms that have long dominated their industry:
When a member of the small business office educates the program manager that
setting aside work for small or small, disadvantaged businesses builds capable suppliers,
a good industrial base, and benefits competition, that program manager is more likely to turn to
his colleagues than to the contractor for trusted advice and direction. That program
manager will be better capable, moreover, of explaining the acquisition process's guiding
principles to the end user. Coordination through awareness-raising and active listening, as
well as even more fundamental factors such as trust, can bring true empowerment to the
acquisition team. Once a sense of trust and common goals is fostered, the diverse
membership of the acquisition team can become something of a "collective ombudsman"
to ensure effective competition as envisioned by FASA.91
The ideals that have emerged from over four decades of FAR—primarily, specific
contracts that minimize the risk of overrun, promote competition in a marketplace, and also
support small businesses—serve as benchmarks for defense innovation programs. Congress has
limited authority over the enforcement and actualization of FAR, while federal agencies and their
individual procurement organs are responsible for making FAR recommendations and rules a
reality.
Thornton, Fine-tuning acquisition reform's favorite procurement vehicle, the indefinite delivery 91
contract. 425.
!35
Fisher
Rapid Acquisition Task Forces Within the DoD
Circumventing the Department of Defense’s traditional acquisition processes, rapid
acquisition outfits are able to contract and receive new products under expedited time frames.
Their presence, however, is most salient during periods of war; these authorities equip
warfighters contingent on immediate battlefield needs. Additionally, they typically accelerate
procurement from Beltway defense vendors, not from smaller firms. Comment on rapid
acquisition entities often concerns the stovepiped nature of these authorities, which isolate
purchases within DoD organizations. Ultimately, the siloing of rapid acquisition authorities
renders them an unreliable conduit for technology startups seeking defense contracts, because
these firms struggle to independently build inroads to these discrete and small procurement
authorities.
I. Rapid Acquisition Favors the Establishment
Jonathan Wong, a RAND Corporation analyst, finds Operation Iraqi Freedom as a
watershed moment for rapid acquisition forces. “Every time the Marine Corps sent me back to
Operation Iraqi Freedom, new and better equipment awaited: radios, armored vehicles, electronic
jammers to foil roadside bombs,” he reminisces, and describes these rapid deployments as a
crucial resource in a theatre of unpredictable terrain and insurgent strategies. The electronic 92
Wong, Jonathan. 2016. Don’t Learn the Wrong Lessons from Rapid Acquisition. Defense One (June 92
23).
!36
Fisher
jammers came from the defense contractor ITT. Oshkosh Defense, a vehicle contractor well-93
known within DoD, won the bid for mine-resistant, ambush-protected (MRAP) vehicles. 94
Following Operation Iraqi Freedom’s affinity towards rapid acquisition authorities, the
Pentagon “enshrined rapid acquisition by including a dedicated section on it in the latest
regulations governing acquisition,” according to Wong Even then, this excitement primarily 95
applied to wares produced by large defense contractors. “The Air Force recently announced that
it is procuring its new B-21 bomber through its rapid capabilities office, and the Navy is setting
up a similar office to speed up acquisitions,” Wong said of these new rapid acquisition efforts. 96
DoD agencies use rapid acquisition authorities to quickly procure Beltway contractor
goods in time-sensitive situations, bypassing the overhead of traditional contracting vehicles like
those in DoD Instruction 5000.02. These authorities, however, rarely purchase of goods from 97
small defense startups, nor do they prioritize this type of vendor. Though rapid acquisition
organizations have the capability to quickly procure new tools, they do not play an intentional
role in the defense innovation ecosystem.
II. Siloing Expedited Procurement Hampers New Technology Adoption
Rapid acquisition authorities are also impractical for the widespread adoption of new
defense technologies. Over 30 rapid acquisition organizations exist, and operate only within their
Shachman, Noah. 2011. The Secret History of Iraq's Invisible War. Wired. June 14.93
Feickert, Andrew. 2011. Mine-Resistant, Ambush-Protected (MRAP) Vehicles: Background and Issues 94
for Congress. Washington, D.C.: Congressional Research Service. Wong, Jonathan. 2016. Don’t Learn the Wrong Lessons from Rapid Acquisition.95
Wong, Jonathan. 2016. Don’t Learn the Wrong Lessons from Rapid Acquisition.96
Department of Defense Instruction 5000.02.97
!37
Fisher
parent defense agencies. (A list of these rapid acquisition organizations can be found at 98
Appendix B). As a result, requests for similar tooling happen independently and become
duplicative processes, as if each contract were a stovepipe feeding just one of many proximate
stoves. In July 2016, the Joint Improvised-Threat Defeat Organization (JIDO) asked Congress
for a $20 million budget to counter unmanned aerial systems (UAS). JIDO was hardly the only 99
DoD agency working on this issue. The Marines also sought counter-UAS technology, as did the
Army Rapid Equipment Forces. If each agency’s rapid acquisition outfit independently 100
pursued counter-UAS products, their efforts would not only be wasteful, but also ignore
opportunities for collaboration and experimentation among multiple DoD agencies seeking
technologies with similar specifications.
Rapid acquisition authorities, and their tendency to prefer well-known contractors and
stovepipe procurement, do not readily facilitate the inclusion of defense startups nor the
expansion of their products among DoD customers.
A Summary of Continuing Challenges in the Defense Innovation Ecosystem
Improving federal acquisition processes, especially in the defense and intelligence
communities, has been a longstanding goal for the US government. Multiple branches of
government, from Congress to the DoD and CIA, have taken action towards the advancement of
novel defense technologies. Dedicated organizations, such as GVCs and work investment
Solis, William. 2011. DOD's Urgent Needs Processes Need a More Comprehensive Approach and 98
Evaluation for Potential Consolidation. Washington, D.C.: Government Accountability Office, GAO-11-273.
Judson, Jen. 2016. Pentagon Asks for More Money to Counter ISIS drones. DefenseNews, July 8, 2016.99
Judson, Jen. 2016. Pentagon Asks for More Money to Counter ISIS drones.100
!38
Fisher
programs, encourage the development of defensive technologies through funding promising
companies and research. FARs serve as benchmarks for rapidly acquiring newer technologies
offered by smaller contractors, and rapid acquisition authorities are the entities empowered by
this discretionary authority. Among these four vehicles for defense innovation, one link between
the federal government and defense startups appears chronically under-serviced: an interface for
connecting defense customers to new products, and especially those from commercially
promising companies previously supported by In-Q-Tel and DARPA. Among GVCs, work
investment programs, FARs geared towards supporting small businesses, and rapid acquisition
organizations, an unmet goal is the rapid procurement of novel tools from small technology
firms.
DIUx’s mission—to bridge DoD entities and companies working on cutting-edge
technology and devise new practices for federal acquisition—endeavors to fill this gap. As DIUx
attempts to interface innovative technology firms with DoD customers, it must navigate the
challenges faced by preexisting interfacing organizations such as the In-Q-Tel Interface Center,
in addition to the issues that prompted the announcement of DIUx 2.0. Through case studies of
DIUx contracts, and examining the perspectives of leadership and procurement personnel from
DIUx and their contractors, this thesis seeks to add nuance to continuing challenge of
streamlining procurement between the defense community and companies at the forefront of
defense technology.
!39
Fisher
!40
Fisher
Chapter III | DIUx’s First Year, its Early Challenges, and its Reinvention
Defense Innovation Unit Experimental (DIUx), announced by Secretary of Defense
Ashton Carter in April of 2015, seeks to address the defense community’s persistent struggle to
procure novel technologies for warfighters and other personnel. As highlighted in the previous
chapter, the defense community has supported R&D efforts with the potential to support DoD
strategy and missions, but the translation of R&D successes into tangible products for the
defense community remains infrequent.
With the creation of DIUx, Secretary Carter affirmed his commitment to bridging this
procurement gap. In his April 2015 Drell Lecture at Stanford University, titled “Rewiring the
Pentagon: Charting a New Path on Innovation and Cybersecurity”, Carter stated, “the DoD, other
federal agencies and tech companies helped to ignite the spark [of cutting-edge technologies],
but Silicon Valley companies nurtured the flame and created unimaginable applications.” 101
Outlining the issue space targeted by a nascent DIUx, he continued, “[…] startups are the leading
edge of commercial innovation,” he said, “and right now, DoD researchers don’t have enough
promising ways to transition technologies that they come up with to application. I want to fix
that too.” 102
The earliest iteration of this program sought to replicate In-Q-Tel’s conversion rates in
having the IC utilize In-Q-Tel portfolio products. DIUx initially established a pilot project with
the IC venture capital firm to leverage In-Q-Tel portfolio companies as early DIUx partners, with
Carter, Ashton. Rewiring the pentagon: Charting a new path on innovation and cybersecurity. Stanford 101
University, Stanford, CA. 2015. Speech. Carter, Ashton. Rewiring the pentagon: Charting a new path on innovation and cybersecurity.102
!41
Fisher
the intention of linking these companies to potential DoD customers. From this pilot, the first 103
iteration of DIUx—later named “DIUx 1.0”—began to take shape. In response to scrutiny from
lawmakers and private industry, Carter announced a redesigned “DIUx 2.0” a year after DIUx’s
introduction. At the time of writing this thesis, 2.0 remains its current iteration, yielding new
critiques from Congress and the startup community.
To compare DIUx’s shifting management structure and mission, this chapter evaluates the
two versions of DIUx in terms of their leadership structures, federal agency partners, targeted
companies, and procurement strategies. DIUx 1.0 recruited leaders from the defense R&D
sphere; paired with a paucity of DoD funding, this first incarnation of DIUx faced difficulties in
facilitating connections and contracts between the DoD and technology startups.
DIUx 1.0: Fumbling Towards a Clear Mission
I. DIUx 1.0 Management Lacking “Silicon Valley” Traits
The first physical incarnation of DIUx landed on August 2015, next to Moffett Air Field
in Mountain View, California. The program’s first director was George Duchak, formerly
director of the Air Force Research Laboratory Information Directorate, a Defense Advanced
Research Projects Agency program manager, and a technology entrepreneur. Setting up offices in
the Army Reserve Support Center, DIUx wanted to establish what it saw as its entrepreneurial
environment. “When the office is complete, the space will be open and collegial, the director
added, and, like a business incubator, people will work together in a collaborative environment
where all can coordinate and share ideas and opportunities. We’re opening up that space so we
Carter, Ashton. Rewiring the pentagon: Charting a new path on innovation and cybersecurity.103
!42
Fisher
can get a lot of cross-pollination and cross-collaboration among the services,” Duchak said, as
well as among Silicon Valley companies. 104
By installing a former DARPA manager as the director of DIUx, this new project
endeavored to bridge R&D and acquisition under the auspices of a leader with extensive
experience in research and entrepreneurship, but not necessarily in acquisition. Technology
companies are already reluctant to contract with the federal government, given the common
perception that government contracting works at a plodding pace, and the introduction of a
longtime defense R&D official did little to assuage those concerns. “Not everyone in Silicon
Valley is going to want to do business with DoD,” argues Andrew Hunter, a former Pentagon
acquisition official. Getting a contract that could take a year or more “just didn’t translate to
Silicon Valley-speak.” 105
DIUx and its leadership should embrace quicker and potentially riskier contracts in order
to appeal to startups with valuable defensive technologies, argues Herbert Lin, a cybersecurity
research fellow at Stanford University. "The people who do procurement and acquisition in
general at the Pentagon want to follow processes that take all the risk out of it. The whole point
of Silicon Valley is to be risky,” Lin said in an interview with Bloomberg Businessweek. Led 106
by a career DARPA and DoD leader, the first iteration of DIUx failed to espouse the risk-taking
and fast-paced qualities often associated with the technology industry.
Pellerin, Cheryl. "DoD’s Silicon Valley Innovation Experiment Begins." Department of Defense News. 104
October 29, 2015. Print. Syeed, Nafeesa. ”The Pentagon Reboots its Silicon Valley Outpost." Bloomberg Businessweek. June 9, 105
2016. Print. Syeed, Nafeesa. ”The Pentagon Reboots its Silicon Valley Outpost."106
!43
Fisher
II. DIUx 1.0 Still Required Expansion of Defense Agency Partners
Within the first year of its operation, DIUx was able to secure federal participants
including the Air Force, though their level of involvement with DIUx matchmaking efforts is
circumspect. In a February 2016 presentation to the Armed Services Committee Subcommittee
on Emerging Threats and Capabilities, the Department of the Air Force announced that
researchers involved with the Air Force Research Laboratory Rapid Innovation Process agreed to
conduct “quick look" assessments of technologies discovered by the DIUx to determine validity
of further collaborative engagements. The definition of “quick look” was unspecified, 107
signaling an unclear level of commitment in acquiring and using products recommended by
DIUx.
Without clear statements of mission needs put forth by defense agencies, identifying
customer-product matches emerges as a difficult practice. Ambiguous specifications for product
solicitations left DIUx to speculate useful tools for a defense agency, instead of responding to
detailed DoD needs. As a program for bridging defense clients with cutting-edge products, DIUx
1.0 faced the challenge of playing middleman to a client with unclear expectations.
III. DIUx 1.0 Target Companies Too Constrained
Through establishing its first physical office in Silicon Valley, DIUx 1.0 took a pragmatic
approach to fostering connections between the defense industry and small companies. Silicon
Valley, known for its prolific startup community and its historic ties to the defense community, is
home to many vendors for defense contracts. Congressional lawmakers, however, worried that
Department of Air Force Presentation to Armed Services Committee Subcommittee on Emerging 107
Threats and Capabilities. February 24, 2016.
!44
Fisher
DIUx 1.0 unfairly favored West Coast-based companies. The House Armed Services
Committee’s Subcommittee for Emerging Threats and Capabilities explained that it was
“concerned by the pinpoint focus on one geographic region,” urging DIUx to perform outreach 108
to a geographically diverse set of startups across the US.
IV. DIUx 1.0 Strategy Restrained to Performing Assists
Another explanation for DIUx 1.0’s labored progress—resulting in few federal contracts
in its first year of operation—was its lack of readily accessible funds. Duchak noted that DIUx
1.0 had no funds to seed companies with high-potential technologies. Instead, DIUx would be
responsible for connecting labs or DoD offices with small companies that offer useful or
promising products, and the government customer is tasked with seeking more information about
the technology available. If this public-private relationship moves to contract, DIUx’s job is to
help that technology vendor navigate DoD acquisition rules and regulations, Duchak noted. 109
DIUx 1.0 primarily served as a middleman between potential defense community
consumers and small, defensive technology vendors. As a result, the young organization faced
two challenges: first, DIUx had to ensure that defense agencies would commit to procuring
technologies from smaller companies, given that DIUx did not have its own procurement budget.
Second, the program had to encourage smaller companies to pursue federal contracts and provide
these companies with guidance throughout the procurement process.
Department of Air Force Presentation to Armed Services Committee Subcommittee on Emerging 108
Threats and Capabilities. February 24, 2016. Pellerin, Cheryl. "DoD’s Silicon Valley Innovation Experiment Begins." Department of Defense News. 109
October 29 2015. Print.
!45
Fisher
Initially, DIUx had to scrounge for funds from the military services, according to Sonny
Sinha, a former Department of Homeland Security official who had coordinated meetings for
DIUx. "When they successfully vetted an innovative capability and it was time to get the
paperwork started, they had a hard time trying to shake loose funding from the services, because
they didn’t have their own pots," he said in a June 2016 interview. DIUx’s impact was diluted 110
by a decision making process that ultimately hinged on other defense agencies and their own
procurement norms and approaches.
Without a definite budget and the ability to conduct procurement under their own terms,
DIUx was left to guide technology companies through standard federal acquisition regulations.
Lawmakers critiqued how DIUx could not streamline contracts when, at most, they could
demystify FAR for technology startups but not work around them to quicken and simplify
procurement. Without offering alternatives to traditional contracts, the program could not subvert
stereotypes about the difficult and protracted nature of working with the federal government.
Anish Goel, a staff member for the Senate Armed Services Committee, noted that DIUx 1.0
failed to upend or develop upon the FAR that discouraged technology companies from working
with the defense community. In a September 2016 panel discussion, Goel claimed,
The original concept of DIUx was to find companies who were doing sort of
groundbreaking technology and marry them up with other organizations in the DoD who
need that sort of technology. But from our perspective the problem before was not that
these companies didn’t know what the needs were in DoD; it’s that they didn’t want to
work with them because of all of the rules that go along with working with DoD. So
Syeed, Nafeesa. "Can the Pentagon Learn to be Flexible?" Bloomberg Businessweek. June 8, 2016. 110
Print.
!46
Fisher
unless you fix all that other stuff in terms of acquisition and contracting...How is DIUx
really contributing to that problem? Having a dating service really is not the problem in
our mind. 111
An April 2016 proposal from the House Armed Services Committee’s Subcommittee for
Emerging Threats and Capabilities echoes Goel’s concern about DIUx’s lack of teeth in directly
addressing FAR obstructions:
The committee is concerned that outreach is proceeding without sufficient
attention being paid to breaking down the barriers that have traditionally prevented
nontraditional contractors from supporting defense needs, like lengthy contracting
processes and the inability to transition technologies. 112
Without offering alternatives to traditional contracting methods, DIUx 1.0’s progress did not
yield a large number of contracts from matches between defense agencies and small technology
outfits. Following DIUx’s first eight months of operations, only three of 20 procurement
opportunities in its pipeline resulted in contracts, prompting a Bloomberg Businessweek
journalist to call this amount of time an “eternity in a region where a successful venture capital
pitch can generate almost instant funding.” Promising technology vendors, including the 113
autonomous indoor drone manufacturer Shield AI and the cybersecurity company Bromium,
expressed interest in working with DIUx. Ultimately, DIUx and these firms found themselves
Ehlinger, Samantha. "Senate Staffer: Lawmakers Still Skeptical of DIUx." Fedscoop. October 5, 111
2016.Print. United States 114th Congress. National Defense Authorization Act for Fiscal Year 2017. Washington 112
DC:, 2016. Print. 2395. Syeed, Nafeesa. "Can the Pentagon Learn to be Flexible?" 113
!47
Fisher
corralled into the typical 18-month DoD contracting timeframe, and failed to scope and sign
contracts as those companies sought to focus on product development instead. 114
Though DIUx 1.0 faced leadership and strategy challenges as a nascent program, their
work still produced some successes. Bob Goodson, the founder and CEO of Quid, waited a year
for the Pentagon to complete each of its first three contracts with his data science and
visualization company. In 2016, Quid pitched a proposal to DIUx in March, and by May, the Air
Force was using its analysis software. "Going through DIUx has been dramatically faster than we
have experienced in the past," Goodson said in an interview. "One thing I found quite notable is
how involved Secretary Carter has been,” he continued. Goodson’s appreciation of Secretary 115
Carter’s involvement in DIUx suggests that, despite general anathema towards the slow pace of
government, the Secretary of Defense’s earnest effort to reach out to technology startups stoked
enthusiasm towards partnering with a government agency—one publicly working towards
efficiency and technical sophistication.
Introducing DIUx 2.0: A Movement Towards Mission Success
While DIUx 1.0 reaped some modest successes, like facilitating a contract between Quid
and the Air Force in two months, Congressional lawmakers expressed concerns about the
program’s focus on matchmaking and its focused presence in Silicon Valley. When DIUx
requested a $45 million budget in the 2017 National Defense Authorization Act (NDAA), the
House included a provision that would withhold 20% of funding authorized until Defense
Secretary Ash Carter submitted a report to Congress about the effectiveness of the new unit and
Kaplan, Fred. 2016. The Pentagon’s Innovation Experiment. MIT Technology Review (December 19).114
Syeed, Nafeesa. "Can the Pentagon Learn to be Flexible?" 115
!48
Fisher
the process for coordinating it with similar defense entities, like DARPA and In-Q-Tel. 116
Congressional pressure, in addition to feedback from the private industry, prompted Secretary
Carter to announce the creation of DIUx 2.0.
On May 11, 2016, Secretary Carter returned to Silicon Valley and announced the launch
of DIUx 2.0 in the program’s Mountain View offices. Striking a conciliatory tone with DIUx’s
critics, Carter emphasized the incorporation of public and private sector feedback on the second
iteration of his ambitious program. Presenting DIUx 2.0 as the next version of a project—as 117
many technology companies do as they finesse their services and products—Carter underscored
DIUx’s mission to empathize with technology startups and their drive for agile development and
resolution.
The next iteration of DIUx featured an overhaul of its leadership in favor of managers
with acquisition experience. Additionally, DIUx 2.0 sought to move beyond matchmaking
between defense agencies and technology vendors, and expand the program beyond Silicon
Valley.
I. DIUx 2.0 Management Folds in Veteran Entrepreneurs
The second version of DIUx brought in an entirely new leadership team, inspired by
partner-based management structures common among venture capital firms. In his speech
announcing 2.0, Carter claimed, “we’re taking yet another page from the Silicon Valley
playbook, making leadership structure at DIUx as flat as any company here.” Carter’s newly 118
United States 114th Congress. National Defense Authorization Act for Fiscal Year 2017. Washington 116
DC:, 2016. Print. 2395. Section 217. Carter, Ashton. "Remarks Announcing DIUx 2.0." Department of Defense News. May 11, 2016.117
Carter, Ashton. "Remarks Announcing DIUx 2.0."118
!49
Fisher
appointed managing partner—and functional director—of DIUx was Raj Shah, a National
Guardsman and F-16 pilot, a combat veteran, and co-founder and CEO of his own cybersecurity
startup. Additional DIUx 2.0 partners included Isaac Taylor, former Head of Operations for
Google’s experimental X research unit; Vishaal Hariprasad, Air Force Reserve Captain, combat
veteran, Bronze Star recipient, and cofounder of a cybersecurity startup, and former Head of
Threat Intelligence at Palo Alto Networks; and Chris Kirchhoff, former civilian adviser to the
Chairman of the Joint Chiefs of Staff, lead author of the White House's Big Data Report, and
former Director of Strategic Planning at the National Security Council. 119
With a slate of partners now at the helm of DIUx, the second iteration of this program
sought to stack DIUx personnel with former public servants and service members bearing private
industry experience. Replacing a career DoD director with a slate of former entrepreneurs and
technology industry employees, DIUx reacted to concerns about bureaucratic leadership. A new
board of leaders could bolster the organization’s standing among small technology outfits wary
of stereotypically slow and out-of-touch government clients.
II. DIUx 2.0 Still Required Expansion of Defense Agency Partners
The introduction of DIUx 2.0 refocused efforts to find defense agency partners willing to
work with DIUx—a partnership in engaging new technology companies and streamlining the
procurement of their products. With the announcement of the new DIUx, Secretary Carter
introduced three measures to prompt defense agencies to take advantage of the program: placing
DIUx under the auspices of the Office of the Secretary of Defense (OSD), defining specific roles
Carter, Ashton. "Remarks Announcing DIUx 2.0."119
!50
Fisher
within DIUx for defense agency partners to adopt and pursue, and recruiting Military Leads to
strengthen communications between defense agencies and DIUx. 120
DIUx 2.0 reported directly to OSD, a shift in the command chain that highlights
Secretary Carter’s investment in DIUx, and highlights this program to defense agencies and
service members. By heightening the profile of DIUx within the DoD, Carter attempted to elicit
enthusiasm for DIUx and its mission among defense agencies, potentially converting these
agencies to DIUx-supported customers of new technology. Carter also gained more day-to-day
jurisdiction over DIUx in this new arrangement, allowing him and future Secretaries of Defense
to closely examine the progress of this innovation program. 121
Carter also hired new Military Leads within DIUx, tasked with better representing
warfighters requesting new technologies and translating defense agency needs into clear and
cogent announcements for technology companies. By explicitly dedicating DIUx personnel to 122
work on the longstanding challenges associated with translating military needs into product
requests private industry companies can fulfill, DIUx 2.0 seeks to improve its middleman
relationship between the defense community and technology entrepreneurs. The presence of
dedicated Military Leads could persuade defense agencies to use DIUx, which would accelerate
contracts while handling overhead on specifying warfighter needs and requesting business
proposals.
III. DIUx 2.0 Target Companies More Diverse
Carter, Ashton. "Remarks Announcing DIUx 2.0."120
Carter, Ashton. "Remarks Announcing DIUx 2.0."121
Carter, Ashton. "Remarks Announcing DIUx 2.0."122
!51
Fisher
“[A] new feature [of 2.0] is that DIUx will be a nationwide release—we're not just
iterating, we're scaling,” Secretary Carter said during his announcement of DIUx’s next
incarnation. DIUx opened a new Cambridge, Massachusetts office in July 2016, seeking to
engage Boston’s biotechnology and bioscience industry. In September 2016, DIUx also 123
opened a new office in Austin, Texas, another American hub of entrepreneurship. 124
Through opening offices beyond Silicon Valley, DIUx 2.0 attempted to address
Congressional concerns that Carter’s new program favored a specific region of the United States.
Additionally, DIUx’s expansion indicates the program’s interest in expanding its contracting
outreach to nontraditional defense contracting industries, such as those related to bioengineering.
To demonstrate success in its “scaling” efforts, DIUx must successfully negotiate contracts to
deploy products developed beyond Silicon Valley, otherwise the expansion of their offices will
not have capitalized on innovation hubs with characteristic specialities, like Boston and its
biotechnology sector.
IV. DIUx 2.0 Strategy Utilizes Nontradtional Contracting Vehicles
With the availability of a Congressionally managed budget, DIUx was able to conduct its
own contracting instead of providing guidance in an otherwise unchanged procurement
process. DIUx 2.0, attempting to move beyond the middleman model some of its critics jested 125
as a “dating service,” introduced a streamlined procurement vehicle: the Commercial Solutions
Opening (CSO), followed by a contract negotiated under a Other Transaction Authority (OTA).
DoD Press Operations. "Secretary Carter Announces DIUx Presence in Austin, Texas." Department of 123
Defense News. July 26, 2016. Print. DoD Press Operations. "Secretary Carter Announces DIUx Presence in Austin, Texas."124
Carter, Ashton. "Remarks Announcing DIUx 2.0."125
!52
Fisher
This combination of solicitation and procurement processes allowed for the focused and
expedited purchase of a vendor’s products—a substantial movement away from the long-term
contracts typically favored by the DoD, as well as R&D relationships built by DARPA
challenges and grants.
The CSO, much like a DARPA Agency Announcement, broadcasts DIUx’s current
technology areas of interest. If a company has the potential to service this defense interest, it can
submit a solution brief. DIUx grants priority to smaller technology companies seeking contracts
with the DoD, and states that they will get back to a company about possible defense client
matches within 30 days. Additionally, DIUx states that the statutory authority behind the CSO
allows DIUx to mirror the contracting practices that commercial companies normally use,
enabling DIUx to design projects, and negotiate payment milestones, terms and conditions, and
intellectual property rights all within a target period of 60 days. 126
These quick contract turnarounds are the result of DIUx’s use of OTAs. OTAs are defined
by what they are not: they are not standard procurement contracts, grants, or cooperative
agreements. Consequently, they are typically not subject to the federal laws and regulations that
apply to federal procurement contracts. The OTAs leveraged by DIUx originated in Section 127
815 of the 2016 National Defense Authorization Act. This Section authorized the Director of the
Defense Advanced Research Projects Agency, the Secretary of a military department, or any
other official designated by the Secretary of Defense to carry out certain prototype projects that
are directly relevant to “enhancing the mission effectiveness of military personnel and the
Defense Innovation Unit Experimental. DIUx Reports First Quarterly Results. Mountain View, CA: 126
United States Department of Defense, October 17, 2016. 1. Air Force Office of Transformational Innovation. 2015. Other transaction authority (OTA) overview. 127
Washington, DC: United States Air Force.
!53
Fisher
supporting platforms, systems, components, or materials proposed to be acquired or developed
by DoD.” In effect, DIUx, when officially designated by the Secretary of Defense, could now 128
contract prototype products from small technology firms typically beyond the DoD’s
procurement orbit. With this new definition of an OTA, an authorized DIUx official had the
ability to purchase prototype projects without having to endure layers of Pentagon bureaucracy.
Additionally, while Section 815 limited contracts under OTAs to a maximum of $250 million,
this upper bound far exceeds the budget DIUx requested for its contracts. Section 815 granted
DIUx the jurisdiction to cull traditional procurement processes in favor of a rapid approach to
testing and implementing promising prototypes for DoD workflows. The discovery of this
particular component of the NDAA was groundbreaking to DIUx 2.0’s strategy: Chris Kirchhoff,
a partner at DIUx, described the discovery of Section 815 “like Thomas Jefferson taking out his
pen and writing the Declaration of Independence.” 129
With the combination of CSOs and OTAs, DIUx 2.0 was able to establish a contract
signing timeframe of 90 days, inclusive of initial outreach to signatures on paper. The 130
unveiling of this schedule was an exciting development for technology vendors, whose previous
attempts to contract with the DoD previously failed, mired in slow timeframes that exceeded the
pace of a company’s product development. Beyond the signing period, the CSO also presents a
purchasing and deployment schedule of at most 24 months in an attempt to mitigate the risk of
schedule or budget overruns. 131
United States Congress, 114th Session. National Defense Authorization Act for Fiscal Year 2016. 128
Washington DC. Section 815. Kaplan, Fred. The Pentagon’s Innovation Experiment.129
Defense Innovation Unit Experimental. DIUx Commercial Solutions Opening Whitepaper. Mountain 130
View, CA: United States Department of Defense, 2016. Print. 4. Defense Innovation Unit Experimental. DIUx Commercial Solutions Opening Whitepaper. 4.131
!54
Fisher
Following the introduction of DIUx’s expedited procurement vehicles and the
confirmation of a DIUx budget, the program awarded twelve agreements in the fourth quarter of
fiscal year 2016, for a total of $36.3 million dollars. Contracts spanned a range of companies in
defensive cybersecurity, unmanned vehicles, and textual and behavioral analysis, with awards
ranging between $400,000 and $12.7 million dollars. These contracts even included Bromium
and Shield AI, two companies whose product pitches previously failed to become contracts under
DIUx 1.0. Appendix C is a list of these FY 2016 Q4 contracts and their award amounts.
A Summary of DIUx’s Evolution and a Question: What Lies Beyond the OTA?
Thus far, DIUx has actively moved to bridge the under-addressed gap between the
defense community and young and innovative technology vendors. Additionally, the transition
from DIUx 1.0 to 2.0 demonstrated a rapid-response approach to critiques from both the public
and private sectors. With overhauls in leadership, strategy, and outreach efforts, DIUx 2.0
explicitly addressed initial concerns about the organization’s stature in a bureaucracy, in addition
to its ability to actually expedite contracts. Through DIUx 2.0 has secured a first round of
contracts, the success of these contracts—evaluated as the procession from bookings to actual
payments and product deployments—remains a crucial criterion for evaluating DIUx 2.0’s ability
to meaningfully advance defense procurement.
The remainder of this thesis investigates challenges within procurement that can only
emerge from a first round of contracts. Indeed, after federal contracts are signed, detrimental
contract term changes or contract breaches are among the many issues that afflict federal
procurement. Through interviews with DIUx personnel and employees at technology startups
!55
Fisher
with DIUx negotiations or contracts, I will examine the booking-to-deployment process
facilitated by DIUx, and assess the circumstances surrounding OTA negotiation and contract
completion. DIUx, in its two years of existence, has made substantial leeway in developing trust
and enthusiasm among technology startups and organizing deals with highly agile companies. In
its current form, will DIUx be able to maintain a productive relationship with technology firms
as they deploy products and seek to expand their customer base?
!56
Fisher
Chapter IV | Case Studies on DIUx 2.0
Following DIUx’s first round of contracts in Q4 2016, companies with DIUx-facilitated
contracts are now in their two-year product deployment stage. To gain a better understanding of
the successes and limitations revealed by this first round of OTA-based procurement, this author
conducted interviews with members of DIUx and technology startups that were involved in the
drafting and execution of these contracts. These interviews, performed over February 2017 to
May 2017, revealed that initial OTAs were effective in placing new technologies in the hands of
American service people. These short-term contracts, however, produced questions about the
continuation of these public-private partnerships beyond the first OTA.
The following three case studies, involving perspectives from DIUx-contracted startups,
the reflections of DIUx personnel, and my personal observations at the drafting of a potential
CSO, examine these concerns about the lifespan of DIUx-fostered relationships. Ultimately,
OTAs emerge as a highly useful contracting vehicle, but they are most effective in quickening
the sale of defense tools. Their short-term timeframe, though good for warfighter in demand of
new technologies and useful for companies seeking to maximize sales, does not guarantee
defense startups a long-term relationship with additional DoD customers. Additionally, DIUx’s
reliance on this tool may stoke tensions over the quantity of contracts versus their quality: should
DIUx maximize OTAs performed or, at the expense of contract count, prioritize the production
of contracts conducive to longer-term engagement? Who among DIUx, DoD agencies, or
technology firms is most responsible for the sustainability of this public-private partnership after
the first OTA?
!57
Fisher
Perspectives from Two DIUx-Contracted Firms
Small defense firms contracting to DIUx encompass producers of hardware and software;
this section evaluates a hardware producer and a software startup’s successes and challenges in
working with DIUx. Software firms face a unique challenge in that their products are highly
malleable and prone to change, introducing challenges in terms of vending to the government
and fulfilling immutable contract specifications while continuing agile product development.
Despite these challenges, both companies express an interest in using DIUx as a starting point for
expanding their relationships with other potential customers in the defense community.
I. Shield AI
Shield AI gives unmanned aerial systems (UAS) additional hardware and software to
perform live image recognition. Such improvements allow UAS to more accurately perceive
their environment and forward appropriate information to warfighters, addressing geospatial
intelligence deficits in high-risk situations. Through DIUx, Shield AI won a $1 million contract
with the Department of Defense to deploy drones equipped with their hardware sensors and
corresponding software.
Brandon Tseng, a former Navy SEAL, co-founded Shield AI and currently works as its
CFO. In his position, he closely worked with DIUx to negotiate and execute upon its OTA.
Before engaging DoD customers through DIUx, Shield AI responded to a number of solicitations
from DARPA and the Rapid Innovation Fund for the DoD, in addition to more traditional
solicitations managed by general Federal business operations output. He found their solicitation
!58
Fisher
process as “slow, archaic, and ill-conceived” given the minimal customer interaction time they
provided. Lacking extensive discussion with a DoD client, Shield AI did not have the
information necessary to create a focused product pitch, let alone draft and negotiate a contract.
Similarly, DARPA solicitations were “not great about describing the product they really want.”
Tseng found DIUx as a useful middleman between consumer and vendor, whose experienced
personnel in both the defense sector and the technology industry facilitated Shield AI’s
communications with DoD purchasers.
The contract Shield AI performed through DIUx even exceeded the payout Shield AI
expected; for the company, a good business to business sales projection was $100,000 in sales
negotiated over the course of 60 days. Through DIUx, Shield AI received four times that amount
within 60 days. The use of OTAs allowed DIUx fulfill contract quantities quickly, and though
these contract costs were extremely modest relative to overall DoD spending, this income is
invaluable for young startups.
Tseng also noted that DIUx granted Shield AI a large degree of autonomy in terms of
executing a contract, and did not micromanage the deployment of the company’s product to the
DoD. He described DIUx’s approach as “broker[ing] a contract, mov[ing] money, but letting you
[the company] lead the program sold.” Through OTAs, DIUx had fashioned a role as a
matchmaker with the power to push contracts efficiently, granting companies and their customers
responsibility for the ultimate outcome of the the contract. Additionally, concerns over the
dilution of company equity—scenarios common to relationships with DARPA or In-Q-Tel—are
not at all present in the OTA procurement process.
!59
Fisher
In this hands-off role, DIUx acts not as an incubator, but acts as a first screen for a
company before it can initiate a relationship with the DoD and enter the defense procurement
process. Though this strategy empowers technology companies to directly work with defense
clients, it also imposes an outsize authority on technology firms to independently develop future
relationships with the DoD after a successful first contract. Without DIUx further cultivating
these business relationships, this expansion can be a difficult endeavor for a startup to pursue
alone, given the DoD’s preexisting close relationships with larger Beltway contractors. Shield AI
maintains a long-term goal of becoming a billion dollar organization over the next ten years,
partially contingent on the goal of becoming an integrated UAS intelligence system for various
federal agencies. While a relationship with DIUx has bolstered Shield AI’s credibility among
United States Government customers, and even drawn the interest of venture capital firms like
Andressen Horowitz, the startup now seems responsible for furthering inroads among federal
clients. Though Shield AI now has cache among one defense agency, they must continue to
cultivate their relationship with other bureaucratic USG consumers. Given their autonomy in
contract execution, whether Shield AI can expect DIUx assistance beyond their first OTA is
unclear.
II. Software Vendor
An off-the-record interview of a federal contracting manager at a software startup
unearthed similar concerns about the longevity and expansion of DIUx-initiated relationships
with defense customers. This firm found DIUx extremely useful in terms of bridging defense
!60
Fisher
clients and nontraditional defense vendors. For the remainder of this section, this interviewee
will be referred to as “Emma”.
“What struck me is that DIUx includes good people, granted their cycles and bandwidth
are extremely tight. [Our contact at DIUx] was very receptive to understanding how we do
commercial business. Our preferred terms, master services agreement structure, baking in things
important to us were more easily expressed through DIUx,” Emma noted. DIUx’s role as an
interlocutor assisted with the drafting and execution of an OTA, but also introduced losses in
translation and a level of overhead that the startup had to surmount in order to advance towards a
contract. Emma described overcoming the challenges of negotiating with a middleman:
There are definitely some frictions in having an intermediate party negotiate on
your behalf, and try and realize the requirements of the government third party they are
basically midwifing. To marry these things together introduces a lot of complexity.
Communication becomes paramount, and making sure that all three parties are aligned is
difficult, especially with a software product. Some points in conversation where we
needed to step back and communicate from first principles perspective to be clear about
what our desired outcome was.
Clear communication is especially crucial for small software contractors, whose products
tend to develop so rapidly that, by the time they are deployed, there is a possibility that their
latest release of a product no longer meets the exact specifications laid out in an OTA. Emma
notes that her company was concerned about being beholden to an OTA’s specifications,
especially because competitors could be advancing their products during that period. The
unpredictability of a software engineering team’s roadmap, rooted a competitive interest to
!61
Fisher
frequently iterate on a product, introduces the notion of “scope creep,” in which a product’s
specifications at one point in time may diverge from the finalized product a company seeks to
sell. Emma stated that “scope creep is more of an issue in the software space [than in hardware].
A bunch of people in the room apart from the software developer have been trained in agile
development and keep up with the news about other vendors doing other things.” Indeed,
products based in code are far more supple than hardware products that depend on supply chains,
prototyping, and physical production. Having to constrain development in order to meet
government contract specifications can hamper a software company’s ability to develop their
software.
Though the multilateral OTA negotiation introduced some communications challenges,
Emma noted that their product deployment “has gone well with [our customer], and they’re
looking to renew.” Within a year, her company’s product deployment reached a nine out of ten
rating in Technology Readiness—a ratings system that evaluates how well-integrated a new
technology is in defense workflows. Similar to Shield AI’s experience, Emma has found that
DIUx has “limited” involvement in their product’s deployment to their customer, empowering
the vendor and client to manage their deployment independent of DIUx. Emma states that these
limitations are also real to the extent “that their contracting shop is the arbiter of what they can
do or can’t do.” For defense startups, their relationship with DIUx may seem bounded to the
cultivation and negotiation of one-to-one OTAs between a customer and a vendor.
Emma’s company also found that the immutability of OTAs restricted her company’s
ability to expand their customer base within the DoD. Following an unsuccessful effort to change
the client scope of their first OTA, Emma commented,
!62
Fisher
I was under the impression that these contract vehicles could have been flexible
and addressed changes, and been in the scope of expanding customers or prototyping
efforts with existing customer. It’s hard to define exactly where this disconnect was; I
think the primary driver is that, the contractual authority viewed OTA in very similar light
as a [more traditional] software and services contract, and didn’t view it as flexible
structure outside of those bounds. If we were to have been successful in using an OTA to
expand to other defense customers, we would need to get things right a priori. It’s hard to
optimize [an OTA] once things have been set in stone.
This lack of flexibility, though crucial to an OTA’s speed, introduces significant—if not
impassable—barriers to amending OTAs in the interest of further deploying a product beyond its
first DoD patron. Companies like Emma’s must jump through additional OTA negotiations and
signings, a duplicative act, to engage new clients.
Emma also found that, throughout the contracting process, DIUx was responsive to her
company’s feedback and learned from its successes and errors from its first OTAs. Speaking
about DIUx, she stated:
They’ve learned from this empirical experience, and now can’t implement it for
our company but can do so for younger companies. I think people [at DIUx] appreciate
this because it’s a trite point: for a small business trying to do business with the
government, the vehicle becomes extremely important. Being able to write contracts for
new customers within the same scope would be a huge value add.
DIUx’s involvement in the first OTA was valuable, and Emma’s firm still hopes to
leverage DIUx connections to expand their consumer base. Though Emma’s firm is now more
!63
Fisher
prepared to design and execute contracts outside the auspices of DIUx, having successfully
serviced their first OTA, they still find the DIUx matchmaking platform advantageous. While the
inflexibility of OTAs made the expansion of preexisting contracts difficult, the expansion of
potential clients through DIUx’s network could assist Emma’s firm with creating a future OTA
that includes multiple customers in its initial design. “It is really helpful that DIUx has this sole-
source justification to pursue contracts and skip step one of competition. Having to the recreate
DIUx construct beyond DIUx is hard,” Emma said of DIUx’s networking capabilities.
Emma’s startup finds value in DIUx’s matchmaking, a perspective mirroring that of
Shield AI—after executing their first defense contract, these companies scope future
relationships with the DoD, potentially through DIUx. For these firms, DIUx would ideally
extend a successful vendor’s relationship to new clients and expand their customer base within
the DoD. For both Shield AI and Emma’s company, these relationships with the DoD have
strengthened their hand with respect to commercial marketplace sales and their perception
among venture capitalists.
The OTA, which leverages tightly constrained terms to expedite procurement, helps
young firms rapidly deploy their products to government. Vendors through DIUx financially
benefit from this expedited transaction and the opportunity to test their products under the
stressors of defense community users. A substantial value-add for these companies, nonetheless,
is the ability to grow their government user base, while using this relationship with the DoD as a
means of advancing their fundraising abilities and reputation among private industry customers.
The maintenance and extension of these DIUx-built ties are crucial to this goal, and OTAs alone
cannot ensure the longevity of these relationships.
!64
Fisher
Perspectives from DIUx Portfolio Managers and Contracting Facilitators
In drafting a CSO and matching potential vendors to DoD customers, DIUx involves a
multitude of actors, ranging from its partners to portfolio managers and technical advisors. To
produce case studies representative of the solicitation process—from (1) outreach to potential
defense community customers to (2) the drafting of a CSO—this author interviewed stakeholders
in both steps. These interviews indicated that DIUx’s most significant internal challenges relate
to the process of translating DoD needs into manageable solicitations for small defense
companies to pursue. Given DIUx’s elevated profile within the DoD, the organization has
yielded an increasing number of DoD technology requests. Congressional has scrutinized DIUx
by the number of contracts performed, pushing the organization to maximize the number of
CSOs and OTAs fulfilled. A project manager and military lead at DIUx spoke about the
successes and challenges in DIUx’s outreach to DoD agencies, in addition to their identification
of promising defense startups. To understand the communications challenges in drafting CSO
language, the author also interviewed a figure closely involved with the drafting of CSOs. This
off-the-record interview gleaned concerns about DIUx’s emphasis on contract count, which risks
being at odds with clearly defined and orchestrated contracts—ones conducive to the formation
of longer-term relationships between the DoD and small technology firms.
I. Building DIUx Relationships through OTAs
To bridge the technology industry with the DoD’s bureaucracy and domain language,
DIUx relies on service members with a noted interest in entrepreneurship to serve as translators
!65
Fisher
between these two institutions. Steve Butow, a DIUx Military Lead, draws on his experience as a
commander and vice chief of staff in the Air National Guard to assist communications between
defense customers and new technology vendors. Tom Wester, a DIUx Project Manager, similarly
utilizes his experience as an officer in the Navy to relay military needs and technicalities. As
“warfighters in residence,” Wester and Butow manage DIUx’s space portfolio, seeking
opportunities to procure novel space technologies for the DoD and cultivate aerospace startups.
Both Wester and Butow are large proponents of OTAs, and perceive them as a boon to
developing defense companies. Wester finds that this vehicle, designed for lightweight, rapid
contracting, discourages a DoD tendency to dwell over extremely precise military specifications.
Wester also commented that OTAs impose limited time and price scopes, ensuring that DIUx can
initiate multiple contracts for similar products in order to “breed competition” and encourage
defense startups to be “faster and more nimble.” Furthermore, Wester believes that OTAs are
beneficial for small firms because they do not require the company to surrender capital to a
government investor. “Companies love us because what we’re giving them is a contract that is
non-dilutive in their equity. They usually give up 10-20% of their capital to a venture capital
firm,” he stated.
Wester also finds that other DoD agencies have embraced DIUx’s use of OTAs. He
disclosed that, out of all the contracts facilitated by DIUx, the DoD has met DIUx spending at a
roughly three to one ratio, meaning that DIUx’s $20 million in Q4 2016 expenditures amplified
to $70 in DoD agency spending on unconventional defense contractors. The growing popularity
of the OTA, paired with direct participation in this process by defense customers, lays a potential
foundation for DoD agencies to independently utilize OTAs. If DoD agencies are able to conduct
!66
Fisher
their own OTAs, the companies they contract may not receive the broad recognition that a
prominent, industry-supportive organization like DIUx can offer.
Though DIUx is enthusiastic about—if not reliant on—OTAs, they still endeavor to and
perceive themselves as a bond between the DoD and innovative technology firms. Wester sees
DIUx as an organization whose mission extends beyond expediting contracts:
DIUx is matchmaker of sorts, but more than that. We’re a relationship manager
for innovation in the valley. A lot of the relationships that we have with vendors within
the valley are ones that people within the DoD could not develop. This knowledge is not
embedded in the DoD ecosystem. In fact, a lot of the companies we work with are in
stealth mode. We frame the art of what is possible.
Wester’s belief that DIUx “frames the art of what is possible” stands at odds with the
perception of Emma’s software firm, which found that DIUx’s OTA-based contracting shop is
limiting and “the arbiter of what they can do or can’t do.” Though DIUx’s embrace of OTAs has
borne immediate successes for companies and DoD warfighters, private firms do not see this
vehicle alone as a reliable tool for sustaining relationships with the DoD. To uphold its mission
to build durable connections between new defense companies and government customers, DIUx
must address the disconnect between its and its vendors’ perceptions of OTAs; more specifically,
DIUx must examine their ability to not just produce an OTA for a firm, but also effectively
integrate successful small vendors in the DoD procurement ecosystem.
DIUx’s focus on OTAs is likely a result of Congressional pressure, in the form of an
irregular budget already once withheld by the House Armed Services Committee. Butow noted
that one of the largest challenges facing DIUx is ensuring that Congress continues to fund this
!67
Fisher
effort. The organization’s focus on producing OTAs, then, is an effort to prove DIUx’s worth in
quantitative terms that appeal to federal budget-makers. This pressure, however, must be
weighed with the sentiments of the stakeholders directly working with DIUx, lest DIUx’s
emphasis on OTAs leaves its relationship-building mission unfulfilled.
II. Skepticism Towards OTA Reliance
An off-the-record interviewee, who played a substantial role in drafting DIUx CSOs, also
shared reservations towards the fast-paced nature of OTAs. Luke, as this thesis will name him,
argues that DIUx may kowtow to Congressional pressure to the extent that it disadvantages its
own organizational goals.
Borrowing the parlance of technology industry-speak, Luke states that DIUx may
embrace the “hotfix” approach to resolving issues. In software, “Hotfixes” are rapid actions to
temporarily mitigate larger problems until a more substantive solution is available. OTAs are a
highly effective use of FAR to expedite procurement and include more startups in defense
contracting, but they insufficiently address the DoD’s institutionalized affinity towards Beltway
contractors. Additionally, the limited scope of the OTA may not provide the necessary exposure
needed to traverse culture and communication gaps between the defense community and young
defense technology outfits.
Specifically, Luke finds that OTAs risk siloing DoD relationships with defense startups in
a manner consistent with rapid acquisition task forces. He stated that, while DIUx has seen more
to defense innovation than Special Forces rapid acquisition, DIUx has still not thought about
scaling their work across their many portfolios and clients. Luke points to the DoD-wide need for
!68
Fisher
better data analysis as one that has been frequently siloed within defense agencies, forcing
duplicative procurement procedures and preventing the cross-pollination of software
recommendations and best practices. “Over five government agencies have come to DIUx to talk
about data analysis tools, and DIUx needs to try and solve problems to bring people together and
break down stovepipes,” contended Luke. DIUx has succeeded at this once before, drafting a
generalized request for counter-unmanned aerial vehicle technology sought by JIDO, the
Marines, and the Army Rapid Equipment Forces. DIUx was able to advance this general
solicitation with the Marine Corps as its first customer; Luke hoped that DIUx can continue to
identify and act upon multilateral defense community needs, and resist a tendency to silo
contracts.
DIUx is also well-served by a healthy skepticism and selectivity towards the customers
and vendors it chooses to connect, argued Luke. In an effort to appease Congress and maximize
OTA throughput, DIUx may execute contract that match vague DoD concerns with startups that
seek the government as their main consumer—a dangerous precedent that could pair the DoD
with firms who do not offer a product of commercially competitive quality. Luke worried that,
without thorough scrutiny of DoD customer demands and potential vendors, predatory firms can
take advantage of DIUx solicitations. “I’m afraid [technology industry] people know they can
pull wool over the eyes of a general,” he commented. DIUx also frequently hosts military
personnel and leadership to introduce them to defense technology startups in an act that could be
tantamount to “tech tourism” as opposed to careful analysis of a startup, its offerings, and their
ability to serve DoD customers. Congressional pressure to advance OTAs manifests in two
forms: potentially poorer contract quality with unclear defense community expectations—driven
!69
Fisher
by excitement over new tech than careful consideration of agency technology needs—and the
inclusion of potentially predatory companies seeking to become reliant on the DoD as their
primary client.
Luke urged DoD customers and technology vendors to see beyond the excitement of the
DIUx program and the fast-paced nature of OTAs, and exercise prudence while building eye-to-
eye expectations on defense needs and procuring the products that address them. “This being on
the same page is crucial for maintaining a trusting relationship between these two entities and
cannot be sacrificed in the name of incrementing a DIUx contract count,” Luke summarized.
Enthusiasm for the OTA vehicle and its rapid returns may place an emphasis on quantity of
contracts versus the quality of the good procured. Moreover, a focus on OTA execution may
displace DIUx’s focus on cultivating longer-term relationships between DoD customers and
small defense vendors.
Tom Wester acknowledged that, as an “experimental organization, [DIUx] still don’t
know the best processes” for fostering and advancing public-private contracting partnerships.
Going forward, DIUx must grapple with the reality that OTAs alone may not be suitable for the
longevity of these relationships. Ironically, OTAs may detract from DIUx’s stated relationship-
building mission in favor of siloed, though quick, “hotfixes” to new technology procurement.
Observations from the Drafting of a DIUx Problem Statement
Having participated in the drafting of a Problem Statement—the identification of a
defense agency challenge, and the beginning of a solicitation process like a CSO—the author had
the opportunity to examine the challenges of CSO drafting firsthand. As DIUx yields additional
!70
Fisher
technology requests following its elevation to the Office of the Secretary of Defense, the
organization has received increasingly general product requests from agencies seeking to
leverage this new DoD program. On April 26, 2017, this author provided technical feedback on a
working draft of a Defense Information Systems Agency (DISA) request for a automated
endpoint security solution for all DISA-issued mobile phone devices. The author found that this
extremely broad request for technology as indicative of another mounting challenge for DIUx; in
order to move towards OTAs and maintain productive relationships with small defense startups,
the organization must heavily amend or even reject sweeping requests for technologies that these
firms cannot accommodate.
The DISA request sought a multitude of integrated services: mobile device management
that included mobile and endpoint security with functionality across multiple mobile operating
systems, including Google Android, Apple iOS, and Windows Phone software. DISA’s definition
of “endpoint security” did not specify which component(s) of the security cycle they sought to
address: the prevention of attacks, the detection of breaches, the analysis of incidents,
appropriate responses, or combinations of these. While the description for the security solution
prescribes the ability for military personnel to download from smartphone “app” stores without
having the DoD whitelist those applications, it also seeks to ensure server-side security for
servers connected to these threats—two vastly different problems in the computer security world,
bridging individual device security and the security of the entire system to which a device is
connected. DISA’s desired product appeared to be an all-encompassing security solution.
In this author’s technical perspective, DISA’s request not only extends far beyond the
services a small firm can offer, but also seems an extremely ambitious request for any firm to
!71
Fisher
address. Enthusiasm for DIUx may promote the drafting of too many “moonshot” problem
statements like these, which expect that innovative firms can produce panaceas for massive
national security problems. Unless DIUx is able to reject DoD problem statements and encourage
customers to draft tightly defined technology needs, the organization will grapple with the
complexity of wedding demands for wholesale security solutions with another reality: small
technology startups develop focused, discrete products.
Such broad DoD demands also pose challenges for DIUx’s contracting methodology.
DISA’s request resembles a systems-wide product deployment characteristic of traditional
procurement vehicles, not a deployment well-served by expedited variants like the OTA. Parallel
to Luke’s concerns—about DoD customers and startups attempting to use DIUx for the sake of
its existence and the possibility of quick monetary returns—this author also sees a need for
selectivity in choosing potential DoD consumers with focused product requests. DIUx is best
served by marrying clients and vendors who can clearly narrow and articulate their product
requests and offerings, building an informed relationship in the process.
Conclusions
Consolidating constructive criticism from private and public sector stakeholders in DIUx,
three themes emerge about the success and limitations of OTAs.
First, the efficiency of OTAs is a boon to DIUx’s widget delivery mission, but when used
alone, may disadvantage small firms seeking to improve their product and expand their
relationships with DoD agencies. The efficiency of OTAs, and the inflexible language needed to
advance a contract at this speed, particularly challenges software firms, who may have to
!72
Fisher
interrupt development schedules to comply with previously set terms in an OTA. If these firms
ever change software development cycles around defense customers, they must hope that an
expanding relationship with the DoD stems from this tradeoff. Such long-term engagement with
the DoD offers these startups a twofold advantage: these firms can engage additional defense
clients, and a rapport with the DoD builds a small outfit’s cache among private investors and
private industry customers. Two-year procurement vehicles alone cannot guarantee the longevity
of these business ties.
Second, the tightly bound nature of OTA terms risks becoming stovepiped procurement
efforts similar to those of special rapid procurement operations. As a channel between the DoD
and private industry, DIUx has the ability to synthesize distinct agency requests for new
technologies and address them all with a particular product. Given that the current nature of
OTAs is as a one-to-one relationship between client and vendor, DIUx has yet to leverage its
middleman status to dismantle procurement silos within DoD agencies and encourage the cross-
pollination of new technologies and best practices.
Finally, maximizing OTA count for the sake of legitimizing DIUx to Congress risks
producing poorly defined contracts not promotive of longstanding private-public relationships.
As DIUx gains prominence within the DoD and the startup community, this pressure to increase
contract count may allow for broad or poorly defined defense needs to make it to the CSO level,
delivering solicitations that are not reasonable nor lucrative to small defense firms. Conversely,
startups interested in exploiting the DIUx’s affinity towards small tech firms may attempt to
deliver a competitively inviable product to the DoD, cognizant of DIUx’s quantitative contract
goal.
!73
Fisher
Going forward, DIUx must recognize the potentially dangerous comforts of relying on
OTAs as relationship managers, and weigh other FAR or networking opportunities to expand and
deepen the presence of defense startups in the procurement ecosystem. Though the OTA has been
successful for the product-procuring component of DIUx’s mission, the organization cannot sit
on this vehicle’s laurels.
!74
Fisher
Chapter V | Conclusion: Towards DIUx 2.1, Built on the Success of OTAs
To evaluate the utility of DIUx in an already crowded defense innovation ecosystem, this
thesis analyzed this program over three sets of criteria. First, it investigated the shortcomings of
preexisting efforts including DARPA, In-Q-Tel, targeted federal acquisition regulations, and
rapid acquisition authorities. Second, this thesis examined DIUx’s ability to meet its mission
statement of streamlining procurement and bolstering business relationships between the DoD
and defense startups. This second analytic axis covered the evolution of DIUx 1.0 to DIUx 2.0,
and the efficacy of these changes in advancing DIUx’s mission. Finally, this thesis incorporated
case studies about DIUx contracts and processes to identify future challenges facing DIUx,
following its first successfully negotiated procurements.
DIUx indeed addresses a substantial and unfulfilled niche within the defense innovation
environment; the organization recognizes that DARPA and In-Q-Tel serve as incubators for high-
potential defense research or products, but do not deliver large returns on technology.
Furthermore, efforts to directly expedite procurement processes and prioritize small vendors
through FAR are difficult for Congress to enforce, given their status as mere discretionary
authorities. Even when expedited forms of FAR are invoked by a defense agency, these efforts
are isolated within individual defense agencies, prompt duplicative efforts to procure similar
technologies, and are not conducive to the sharing and adoption of new technologies from
nontraditional defense vendors. DIUx’s organizational imperative—to streamline procurement as
a means of procuring new widgets for the DoD and support defensive technology startups—
addresses a gaping procurement gap within the defense community. Between the DoD’s search
!75
Fisher
for cutting-edge technologies and young defense vendors seeking to sell novel tools, these two
institutions did not share a clear interlocutor. These public-private business connections and the
pursuit of procurement contracts remained elusive in the defense innovation ecosystem
preceding DIUx.
Though its mission addressed a crucial need in defense procurement, DIUx’s first
operating year was marked by skepticism from the private industry and Congress; only when the
organization discovered the authorities granted by OTAs did DIUx begin to stride towards its
stated goals. DIUx conducted OTAs between defense customers and defense startups, using the
broad mandate given to this contract vehicle to reduce average contract negotiation to two
months and deployment timeframes to two years. With over twenty contracts orchestrated by
DIUx by the completion of this thesis, the DoD’s ambitious new procurement project has reaped
clear benefits in terms of equipping warfighters with new tools, in addition to facilitating the
sales of small defense firms and their burgeoning relationships with the DoD.
The widespread use of OTAs, however, prompts questions about the long-term futures of
DIUx-negotiated contracts. The short timeframe of OTAs may not be conducive to the
maintenance or expansion of a startup’s business ties with the DoD. This contracting vehicle
does not explicitly support startups seeking to develop the breadth and depth of their working
relationships with the defense community. Additionally, Congressional pressure on DIUx and its
ability to produce contracts may incentivize the agency to focus on contract quantity over
quality. This tension between high contract throughput—a means of legitimizing DIUx under
Congressional review—and the option to dedicate support and expansion opportunities to
preexisting DIUx-endorsed startups poses a organizational dilemma. Entering DIUx’s third year
!76
Fisher
of existence, the DoD agency must now grapple with the best means of continuing their
relationship-building work beyond the facilitation of a first OTA. DIUx 2.0 does not face the
severe private industry and Congressional criticism facing DIUx 1.0, and the organization's use
of OTAs has fulfilled the expedited contracting goals of the organization. Towards DIUx’s
secondary goal of cultivating robust relationships between the DoD and the defense startup
community, the organization must weigh engagement opportunities beyond a first OTA. This
action, nonetheless, does not seem to require an organizational overhaul as the transition from
DIUx 1.0 to 2.0 did, suggesting that the work ahead is indicative of the need for a DIUx 2.1.
Moving onward, the DoD’s defense procurement experiment must grapple with the reality that a
first OTA is not a guarantor of continued public-private ties between the defense world and
technology startups.
Further research into the workings and efficacy of DIUx remains. While this thesis
identifies Congressional pressure as a negative force in DIUx’s ability to work with private
industry expectations, the converse also merits investigation: how feasible are private industry
expectations and demands for DIUx, and how much do they conflict with DoD goals? Indeed, a
startup’s interest in increasing their breadth of defense customers, potentially through multi-
customer OTAs, faces a bureaucracy long hostile to concessions on particular implementation
details and the procurement of general off-the-shelf products. In 1961, the four military
establishments of the Army, Navy, Air Force, and Marine Corps could not come to a consensus
on a common belt buckle, to the ire of then-Secretary of Defense Robert McNamara. For 132
Chicago Tribune Reporting Staff. 1964. Buckles and Beer. The Chicago Tribune. April 1, 1964.132
!77
Fisher
disparate branches of the DoD to agree on the deployment of unified tooling, and for this to
become a common practice, seems like an unlikely prospect.
Apart from identifying tensions introduced by private actors in DIUx processes, future
research can also delve into other authorities vested in the National Defense Authorization Act
that can facilitate the connection of DoD customers and promising young defense vendors. This
study of specific options for the evolution of DIUx 2.1 need not only rely on the NDAA, but also
consider aspects of organizational theory and relationship-building within bureaucratic
institutions.
This further research is pressing, especially as American adversaries utilize new
technologies across the physical and cyber domains and the defense community finds itself in a
commercial arms race. In the midst of World War II and the Cold War, the close relationship
between new defense actors and the military bureaucracy gave rise to an era of American
innovation and the accelerated the development and deployment of cutting-edge technologies. As
the United States faces a new breed of adversary—of an insurgent, decentralized, and
unpredictable form—an embrace of technology and the notion of offset becomes critical once
again. DIUx plays a crucial role in re-igniting this interest in using innovative, if not
prototypical, technologies to distinguish the United States defense community from its
international peers.
!78
Fisher
!79
Fisher
!80
Fisher
References
Written Resources
In-Q-tel form 990. 2013.
Air Force Office of Transformational Innovation. 2015. Other Transaction Authority (OTA) Overview. Washington, DC: United States Air Force.
Army Test and Evaluation Command Forward Operational Assessment Team XVIII. 2012. Forward Operational Assessment: Palantir Operational Assessment Report. Fort Hood, TX: Army Test and Evaluation Command.
Arnold, Scot, and Bruce Harmon. 2013. The Relative Costs and Benefits of Multi-Year Procurement Strategies. Alexandria, VA: Institute for Defense Analyses, NS D-4893.
Ballard, Gen Christopher. 2013. Training Requirements to Maintain Proficiency on Distributed Common Ground System-Army (DCGS-A). Kabul, Afghanistan.
BBC Newswire. 2016. Kunduz Hospital Bombing: 16 US Forces 'Disciplined'. BBC, April 28th, 2016.
Blank, Steve. 2016. Hacking for Defense @ Stanford – Making the World a Safer Place.
Brander, James, Qianqian Du, and Thomas Hellman. 2010. The Effects of Government-Sponsored Venture Capital: International Evidence. National Bureau of Economic Research (November).
Breitzman, Anthony, and Diana Hicks. 2008. An Analysis of Small Business Patents by Industry and Firm Size. Haddonfield, NJ: Small Business Administration.
Brill, Stephen. 2017. Donald Trump, Palantir, and the Crazy Battle to Clean Up a Multibillion-Dollar Military Procurement swamp. Fortune.
Campbell, Gen John. 2015. Department of Defense Press Briefing by General Campbell via Teleconference from Afghanistan. Department of Defense Press Operations, November 25, 2015.
!81
Fisher
Carter, Ashton. 2016. Remarks Announcing DIUx 2.0. Department of Defense News, May 11, 2016.
———. 2015. Rewiring the Pentagon: Charting a New Path on Innovation and Cybersecurity. Stanford University, Stanford, CA.
Chicago Tribune Reporting Staff. 1964. Buckles and Beer. The Chicago Tribune. April 1, 1964.
Christensen, David, David Searle, and Caisse Vickery. 1999. The Impact of the Packard Commission's Recommendations on Reducing Cost Overruns on Defense Acquisition Contracts. Acquisition Review Quarterly (Summer).
Clark, Major, and Chad Moutray. 2004. The Future of Small Businesses in the U.S. Federal Government Marketplace. Journal of Public Procurement 4 (3).
Cooper, Michael. 2002. Department of Defense Implementation of Acquisition Reform Initiatives. Master's Thesis., Naval Postgraduate School.
Czelusniak, Dan, and Philip Rodgers. 1997. The Kaminski Initiative: Our Most Significant Piece of Unfinished Business in Reforming the Defense Acquisition Process. Paper presented at 1997 Acquisition Research Symposium, Naval Postgraduate School, Monterey, CA.
Defense Advanced Research Projects Agency. About DARPA. in DARPA [database online]. Arlington, VA, [cited December 12 2017]. Available from http://www.darpa.mil/about-us/about-darpa.
———. Contract management. in DARPA [database online]. Arlington, VA, [cited December 12 2017]. Available from http://www.darpa.mil/work-with-us/contract-management.
Defense Innovation Unit Experimental. 2016. DIUx Commercial Solutions Opening Whitepaper. Mountain View, CA: United States Department of Defense.
———. 2016. DIUx Reports First Quarterly Results. Mountain View, CA: United States Department of Defense, October 17.
Department of Defense Instruction 5000.02, (2015).
!82
Fisher
DoD Press Operations. 2016. Secretary Carter Announces DIUx Presence in Austin, Texas. Department of Defense News (September 14).
———. 2016. Secretary Carter Opens Second DIUx Location in Boston, Updates DoD Outreach to Tech Community. Department of Defense News (July 26).
Ehlinger, Samantha. 2016. Senate Staffer: Lawmakers Still Skeptical of DIUx. Fedscoop (October 5).
Federal Procurement Data System. Indefinite Delivery Contract. in Federal Procurement Data System [database online]. Washington, DC, 2017]. Available from https://www.fpds.gov/help/Indefinite_Delivery_Contract.htm.
Feickert, Andrew. 2011. Mine-Resistant, Ambush-Protected (MRAP) Vehicles: Background and Issues for Congress. Washington, D.C.: Congressional Research Service.
Felter, Col Joseph. 2017. Solving National Security Issues with the Lean Launchpad. Paper presented at Hacking for Defense and Hacking for Diplomacy Educatiors & Sponsors Course, Georgetown University, Washington, D.C.
Gansler, James, William Lucyshyn, and Michael Arendt. 2009. Competition in Defense Acquisitions. College Park, MD: University of Maryland Center for Public Policy and Private Enterprise.
Government Accountability Office. 2015. Defense Advanced Projects Research Agency: Key Factors Drive Transition of Technologies, but Better Training and Data Dissemination can Increase Success. Washington, DC: Government Accountability Office, GAO-16-5.
Grasso, Valerie. 2000. Defense Acquisition Reform: Status and Current Issues. Washington, DC: Congressional Research Service.
In-Q-Tel. Submit a business plan FAQ. in In-Q-Tel [database online]. Arlington, VA, [cited 11/22 2016]. Available from https://www.iqt.org/business-plan-submission-guidelines/submit-a-business-plan-faq/.
!83
Fisher
Judson, Jen. 2016. Pentagon Asks for More Money to Counter ISIS drones. DefenseNews, July 8, 2016.
Kaplan, Fred. 2016. The Pentagon’s Innovation Experiment. MIT Technology Review (December 19).
Lerner, Josh, and Felda Hardymon. 2005. In-Q-tel: Harvard Business School case 9-804-146.
Lyngaas, Sean. 2016. Army Eyes DCGS Reforms on Capitol Hill. Fcw (June 1).
Mara, Andrew. 2011. Maximizing the Returns of Government Venture Capital Programs. DEFENSE HORIZONS National Defense University(71) (January).
Mitchell, Ellen. 2016. How Silicon Valley's Palantir Wired Washington. Politico, August 14, 2016.
Pellerin, Cheryl. 2015. DoD’s Silicon Valley Innovation Experiment Begins. Department of Defense News, October 29, 2015.
Porter, Michael. 2000. Location, Competition, and Economic Development: Local Clusters in a Global Economy. Economic Development Quarterly 14 (1) (February).
Senate Armed Services Committee, Subcommittee on Emerging Threats and Capabilities, Strategy and Implementation of the Department of Defense's Technology Offsets Initiative in Review of the Defense Authorization Request for Fiscal Year 2017 and the Future Years Defense Program, 114th sess., 2016.
Ritschel, Jonathan. 2012. Efficacy of US Legislation in Military Acquisition Programmes: Nunn–McCurdy Act Unveiled. Economic Papers 31 (4) (December).
Rosenberg, Matthew. 2016. Pentagon Details Chain of Errors in Strike on Afghan Hospital. New York Times, April 29, 2016.
Schwartz, Moshe, and Charles O'Connor. 2016. The nunn-McCurdy act: Background, Analysis, and Issues for Congress. Washington, DC: Congressional Research Service.
!84
Fisher
Shachman, Noah. 2011. The Secret History of Iraq's Invisible War. Wired. June 14.
Solis, William. 2011. DOD's Urgent Needs Processes Need a More Comprehensive Approach and Evaluation for Potential Consolidation. Washington, D.C.: Government Accountability Office, GAO-11-273.
Syeed, Nafeesa. 2016. Can the Pentagon Learn to be Flexible? Bloomberg Businessweek (June 9).
———. 2016. The Pentagon Reboots its Silicon Valley Outpost. Bloomberg Businessweek (June 8).
Thornton, Karen. 2002. Fine-Tuning Acquisition Reform's Favorite Procurement Vehicle, the Indefinite Delivery Contract. Public Contract Law Journal 31 (3) (Spring).
Tucker, Patrick. 2016. The War Over Soon-to-be-Outdated Army Intelligence Systems. Defense One (July 6).
United States Congress, 114th Session. National Defense Authorization Act for Fiscal Year 2016. Washington DC.
United States Congress, 114th Session. National Defense Authorization Act for Fiscal Year 2017. Washington DC.
Van Atta, Richard, Michael Lippitz, Jasper Lupo, Rob Mahoney, and Jack Nunn. 2003. Transformation and Transition: DARPA’s Role in Fostering an Emerging Revolution in Military Affairs. Alexandria, VA: Institute for Defense Analyses.
Vehse, Robert, Stephen Nygren, and Duane Butherus. 1991. Managing an R&D Contract with the Government. AT&T Technical Journal (March/April).
Walker, David. 2016. Department of the Air Force: Presentation to the House Armed Services Committee Subcommittee on Emerging Threats and Capabilities. Washington, D.C..
Wardle, Caroline. 1999. Obtaining Federal Funding. Arlington, VA: National Science Foundation.
!85
Fisher
Warrick, Joby. 2017. Use of Weaponized Drones by ISIS Spurs Terrorism Fears. The Washington Post, February 21, 2017.
Wong, Jonathan. 2016. Don’t Learn the Wrong Lessons from Rapid Acquisition. Defense One (June 23).
Wonglimpiyarat, Jarunee. 2006. The Dynamic Economic Engine at Silicon Valley and US Government Programmes in Financing Innovations. Technovation (26).
Yannuzzi, Rick. 2000. In-Q-tel: A New Partnership Between the CIA and the Private Sector. Defense Intelligence Journal 25 (26).
!86
Fisher
Interviews
Interview with Brandon Tseng, Cofounder and CFO of Shield AI. May 1, 2017. Stanford,
CA.
Interview with Steve Butow, Military Lead at DIUx. May 2, 2017. Stanford, CA.
Interview with Tom Wester, Military Lead at DIUx. May 2, 2017. Stanford, CA.
Interview with "Emma," a federal contracting manager at a DIUx-contracted firm. May 4,
2017. Stanford, CA.
Interview with "Luke," a figure closely involved in DIUx solicitation drafting. May 12,
2017. Stanford, CA.
!87
Fisher
Appendices
Appendix A: Federal Acquisition Regulations, 1971-2013
Ritschel, Efficacy of US Legislation in Military Acquisition Programmes: Nunn–McCurdy Act unveiled. 493.
!88
Fisher
Appendix B: DoD Entities Involved in the Fulfillment of Urgent Needs
Solis, William. 2011. DOD's Urgent Needs Processes Need a More Comprehensive Approach and Evaluation for Potential Consolidation. Washington, D.C.: Government Accountability Office, GAO-11-273.
!89
Fisher
Appendix C: Q4 2016 DIUx Contracts
!90
Fisher
!91
Fisher
!92
Fisher
Defense Innovation Unit Experimental. DIUx Reports First Quarterly Results. Mountain View, CA: United States Department of Defense, October 17, 2016.
!93